US20110040751A1 - Temporal ranking of search results - Google Patents
Temporal ranking of search results Download PDFInfo
- Publication number
- US20110040751A1 US20110040751A1 US12/911,696 US91169610A US2011040751A1 US 20110040751 A1 US20110040751 A1 US 20110040751A1 US 91169610 A US91169610 A US 91169610A US 2011040751 A1 US2011040751 A1 US 2011040751A1
- Authority
- US
- United States
- Prior art keywords
- user
- information
- event
- weight
- search results
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Abandoned
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9537—Spatial or temporal dependent retrieval, e.g. spatiotemporal queries
Definitions
- search engine services such as Google and Yahoo!, provide for searching for information that is accessible via the Internet. These search engine services allow users to search for display pages, such as web pages, that may be of interest to users. After a user submits a search request (i.e., a query) that includes search terms, the search engine service identifies web pages that may be related to those search terms. To quickly identify related web pages, the search engine services may maintain a mapping of keywords to web pages. This mapping may be generated by “crawling” the web (i.e., the World Wide Web) to identify the keywords of each web page. To crawl the web, a search engine service may use a list of root web pages to identify all web pages that are accessible through those root web pages.
- the keywords of any particular web page can be identified using various well-known information retrieval techniques, such as identifying the words of a headline, the words supplied in the metadata of the web page, the words that are highlighted, and so on.
- the search engine service may generate a relevance score to indicate how relevant the information of the web page may be to the search request based on the closeness of each match, web page importance or popularity (e.g., Google's PageRank), and so on.
- the search engine service displays to the user links to those web pages in an order that is based on a ranking determined by their relevance.
- a method and system for ranking search results based on the temporal proximity to the event to which the search results relate is provided.
- An information dissemination system ranks the search results based on a temporal weight assigned to each search result.
- the temporal weight is an indication of the importance to a user that varies with time.
- the information dissemination system calculates a temporal weight that is based on the temporal proximity of the event that is related to the search result. The temporal weight may be used to re-rank the search results.
- FIG. 1 is a block diagram that illustrates the processing of the information dissemination system, according to some embodiments.
- FIG. 2 is a block diagram that illustrates selected components of the information dissemination system, according to some embodiments.
- FIG. 3 is a flow diagram that illustrates the processing of the profiler component, according to some embodiments.
- FIG. 4 is a flow diagram that illustrates the processing of the profiler component to identify key terms, according to some embodiments.
- FIG. 5 is a flow diagram that illustrates the processing of the profiler component to generate a query, according to some embodiments.
- FIG. 6 is a flow diagram that illustrates the processing of the profiler component to incrementally generate a topic, according to some embodiments.
- FIG. 7 is a flow diagram that illustrates the processing of the information garnerer component, according to some embodiments.
- FIG. 8 is a flow diagram that illustrates the processing of the ranker component, according to some embodiments.
- FIG. 9 is a graphical depiction of weight values as a function of proximity to an event, according to some embodiments.
- FIG. 10 is a graphical depiction of weight values as a function of proximity to an event, according to some embodiments.
- FIG. 11 is a flow diagram that illustrates the processing of the ranker component to calculate a temporal weight of a search result, according to some embodiments.
- FIG. 12 is a flow diagram that illustrates the rendering information based on an event class corresponding to a user event, according to some embodiments.
- FIG. 13 is a flow diagram that illustrates the creating of an event class, according to some embodiments.
- FIG. 14 is an example data listing that illustrates an event class, according to some embodiments.
- FIG. 15 is a flow diagram that illustrates processing an event class, according to some embodiments.
- an information dissemination system identifies the interests of a user and adaptively disseminates to the user the information that is relevant to the user's interests.
- the user's interests may be identified from information derived from various sources, such as, by way of example, the user's calendar entries, email messages, World Wide Web (web) usage information, documents, enterprise resource planning (ERP) data, and the like.
- the information dissemination system then creates a profile for the user, and includes in the profile the details about the user's interests.
- the information dissemination system may include in the user's profile information regarding: the information sources to query for information to render to the user; the user events (e.g., meetings, appointments, etc.) which were analyzed to determine the user's interests; the topics derived from the analyzed events; the queries generated for the analyzed events; a language model of terms that are relevant to the user; the user's preferences for receiving the relevant information; and other information regarding the user.
- the topic can be defined as any cohesive concept of the event.
- the topics may include the subject of the conference, the people presenting in the conference, the people attending the conference, the location of the conference, the organizations represented at the conference, travel information for the user (e.g., origination city, destination city, travel mode, flight numbers, hotel reservations, rental car reservations, etc.), and weather information.
- travel information for the user e.g., origination city, destination city, travel mode, flight numbers, hotel reservations, rental car reservations, etc.
- weather information e.g., weather information for the user.
- the terms that are relevant to the user may be derived from the various sources used to identify the user's interests.
- the information dissemination system periodically executes the queries in the user's profile against the appropriate information sources and obtains the search results.
- search result or any variant thereof generally refers to results which are filtered using a query.
- the information dissemination system then ranks the search results according to various factors, and renders the ranked search results to the user.
- the information dissemination system obtains feedback from the user regarding the rendered search results and incorporates the feedback into the user's profile.
- the information dissemination system may provide a user interface (UI) through which the user can view and/or modify the details of the user's interests maintained in the user's profile. For example, a user may use the UI to specify the level of importance and/or relevance of the events, topics, terms, information sources, and other information included the user's profile.
- the information dissemination system may also monitor the user's interaction with the rendered search results and make appropriate modifications to the details of the user's interests maintained in the user's profile.
- the information dissemination system may also periodically update the user's profile with more recent or updated details about the user's interests. In this manner, the information dissemination system, over time, becomes increasingly responsive to the user's information needs.
- the information dissemination system identifies key terms from the terms that are relevant to a user and generates the queries from the key terms.
- the information dissemination system can identify the key terms from the terms extracted from the user's messages and events contained in selected folders.
- the information dissemination system may scan the user's recent email folders (e.g., inbox folder and sent items folder) and extract the terms from these email folders.
- the information dissemination system can then identify the emails in the other folders that are related to the extracted terms, and extract the terms from these related emails.
- the information dissemination system creates clusters of terms that are similar. Each cluster is associated with a set of terms.
- the information dissemination system then identifies events, such as the user's appointments by scanning the entries in a calendaring application, and, for each event, identifies the clusters associated with the event.
- the information dissemination system then creates one or more topics for each event. Each topic may be considered a “container” that contains information that is relevant to its associated event, such as an event name, an event duration, etc.
- the information dissemination system identifies the key terms in the identified clusters and generates a query from the key terms. The information dissemination system then places the generated query into the topic that is associated with the event, and includes the topic in the user's profile.
- the information dissemination system may use various techniques to generate the clusters of terms.
- the information dissemination system uses a K-means algorithm.
- the information dissemination system provides to the algorithm a number K that specifies the desired number of clusters.
- the algorithm takes the first K terms as the centroid of a unique cluster.
- the centroid is the mean position of the cluster (e.g., the mean position of the terms in the cluster).
- the algorithm assigns each of the remaining terms to the cluster with the closest centroid to the term (e.g., the cluster where the difference between the centroid and the term is minimized).
- the algorithm recomputes the cluster centroids based the previous pass and then reassigns each term to the cluster with the centroid that is closest to the term.
- the algorithm may be performed for a fixed number of passes (e.g., three) or until the clustering converges on a solution.
- the information dissemination system may use any of a variety of well-known algorithms, including maximum likelihood estimation, spectral clustering, and so on, to generate the clusters of terms.
- the information dissemination system identifies the key terms in a cluster based on an overall weight assigned to each term.
- the overall weight of a term may be derived from a combination of a type weight of the term and a relevance weight of the term.
- the information dissemination system first filters the terms to reduce the number of terms that may be identified as key terms. For example, the information dissemination system may filter out the terms that are Uniform Resource Locators (URLs), and terms that are made up of numbers in order to reduce the number of candidate key terms. For each of the remaining terms, the information dissemination system can calculate a type weight and a relevance weight.
- the type weight is a weight that is assigned to a term based on the type of the term.
- the information dissemination system may assign a larger type weight to a term that is a name, such as a name of a person, than to a term that is comprised of multiple words.
- a term that is comprised of multiple words may in turn be assigned a larger type weight than a term that is a single word.
- the information dissemination system may also assign different type weights to a term based on whether the term appears as part of the subject text, body text, contacts text, or other parts of the event.
- the relevance weight of a term is a measure of relevance of the term to the user and, in particular, to the user's interests.
- One common technique for determining the relevance of the term to the user is based on term frequency and inverse document frequency (TF*IDF).
- Term frequency refers to the number of occurrences of a term within a document
- the inverse document frequency refers to the inverse of the number of documents that contain that term.
- the documents generally refer to the various information sources used to identify the user's interests.
- the information dissemination system may identify the terms that are over represented and the terms that are underrepresented as terms that are of unlikely relevance to the user. To each of these terms, the information dissemination system assigns a low relevance weight. To each of the remaining terms, the information dissemination system can assign a relevance weight based upon the term's TF*IDF metric.
- the information dissemination system may use any of a variety of other well-known techniques for determining the relevance of a term.
- the information dissemination system may then derive an overall weight for a term based on the term's type weight and relevance weight. For example, the overall weight of a term may be derived by multiplying the type weight and the relevance weight of the term. In some embodiments, the overall weight of a term may be derived from the term's relevance weight. The information dissemination system may then identify as key terms the terms whose overall weights are above a predetermined threshold. The information dissemination system may maintain the terms and their assigned weights in the user's profile, for example, as part of the language model.
- the information dissemination system generates a query from the key terms by using the highest six ranking key terms.
- the information dissemination system first ranks the key terms based on the overall weights assigned to the key terms.
- the information dissemination system then identifies the highest six ranking key terms and generates the query by the following equation:
- the information dissemination system incrementally generates topics that may be of interest to a user.
- the user may provide an indication that some event is newsworthy (i.e., of interest to the user).
- the information dissemination system identifies a user interest.
- the information dissemination system can identify a topic from the specified event as being of interest to the user. Any one of the other topics that are related to the specified event may also be identified as the user interest.
- the information dissemination system then extracts the terms that are associated with the topic and, from these terms, identifies the key terms.
- the information dissemination system can identify the key terms based on the overall weights assigned to the terms, as described above.
- the information dissemination system then generates a query from the key terms and includes the query in the user's profile for execution.
- the information dissemination system may also maintain the extracted terms and their assigned weights in the user's profile, for example, as part of the language model.
- the information dissemination system garners information for a user at periodic intervals. At a periodic interval, the information dissemination system executes the queries against the information sources specified in the user's profile to obtain search results.
- the information sources may include MSN Search, MSN News Search, and other information sources which index blogs, news feeds, web pages, document sources, webcasts, videocasts, audiocasts, etc., and which makes the search results available via an interface.
- the information dissemination system then generates a ranked list of information items by ranking the items of information included in the search results.
- the information dissemination system renders the ranked search results to the user.
- the information dissemination system initializes the periodic intervals to default values, which may be modified by the user.
- the information dissemination system garners information related to events that are within a time range, such as a date range. For example, the information dissemination may initialize the date range to one week (e.g., a period of seven days starting from the day before the current day (today)). In this instance, the information dissemination system garners information by executing the queries generated for the topics corresponding to the events that are within the date range. In some embodiments, the information dissemination system allows the user to initiate the garnering of information. For example, the information dissemination system may provide a UI through which the user can specify a date range of interest and request the garnering of information.
- the information dissemination system ranks the search results based on a total weight assigned to each search result.
- the total weight of a search result may be derived from a combination of a static weight of the search result, an adaptive weight of the search result, and a temporal weight of the search result.
- the static weight may be derived from various weighting factors such as the importance of the topic/event to the user, the importance of the information source, the ranking provided by the information source, and the relevance of the terms to the user's interests.
- a static weight for a search result may be a combination of the importance of the topic that corresponds to the search result, the importance of the information source that produced the search result, the rank provided by the information source that produced the search result, and the relevance of the terms in the search result to the user's interests.
- the relevance of the terms may be determined using the language model included in the user's profile.
- a cosine similarity measure can be employed to measure the similarity of the terms in the search result to the terms in the language model.
- the adaptive weight is a weight that is assigned to a search result based on user preferences and/or feedback. For example, the user may specify additional information sources from which to retrieve information. The user may indicate a preference order for the information sources.
- the user may also indicate a preference for a certain type of search result over another type of search result.
- the information dissemination system may also discern the user's preferences by monitoring the user's interactions with the information dissemination system and, in particular, the information rendered by the information dissemination system.
- the information dissemination system may maintain information regarding the user's preferences and interactions in the user's profile.
- the temporal weight is a weight that is assigned to a search result based on the proximity of the event that corresponds to the search result. Temporal weighting is further discussed below.
- the information dissemination system may then derive a total weight for each search result based on the search result's static weight, adaptive weight, and temporal weight. For example, the total weight of a search result may be derived by adding the static weight, adaptive weight, and temporal weight of the search result.
- the information dissemination system then ranks the search results according to their total weights and renders the ranked list of search results.
- the information dissemination system renders only a predetermined number of search results. For example, the information dissemination system may only render the highest twenty ranking search results.
- the information dissemination system may alert users regarding highly relevant search results.
- the information dissemination system may alert a user by providing an indication on the user's client computing device.
- the information dissemination system may cause a particular icon or indicator to appear on a UI on the user's client computing device.
- the information dissemination system may also vary the color of a displayed icon or indicator to indicate the presence of search results of varying relevance.
- the information dissemination system may send the user a text message informing the user of the presence of relevant search results.
- Temporal ranking is based on the notion that the importance of information varies with time.
- a user may have two events scheduled in the user's calendar: a first event regarding a meeting next week in New York, and a second event regarding the flight to New York for this meeting.
- the importance of the meeting event will be high (i.e., of interest to the user), and any information items (e.g., news) related to the meeting event will also be high.
- the flight event also increases in importance, and information items related to the flight event, such as flight delay information, also increases in importance.
- the flight event will have vastly diminished in importance to the user, and related information items may no longer be of interest to the user.
- the events have importance to the user, and information items related to the events also have importance to the user, but this importance varies according to time and, more specifically, the proximity to the events.
- the information dissemination system ranks the search results based on a temporal weight assigned to each search result.
- the temporal weight is an indication of importance of a search result to a user that varies with time.
- the information dissemination system calculates a temporal weight that is based on the temporal proximity of the event that is related to the search result.
- the value of the temporal weight of a search result may increase (e.g., rise) exponentially in the time period before the start of the related event.
- the value of the temporal weight may remain relatively constant (e.g., flat), or may rise to a peak at a point in time during the related event and decrease thereafter.
- the value of the temporal weight may decrease (e.g., drop) exponentially to zero in a very short amount of time.
- the value of the temporal weight may be calculated using an exponential function of the following form:
- rti represents a time interval (e.g., time from the current time to the start of the event) compared to the length of the event (e.g., T 2 ⁇ T 1 , where T 2 is the end time of the event and T 1 is the start time of the event)
- k and m are weight factors whose values are chosen to vary the shape of the function as desired.
- the values chosen for the weight factors k and m may be different before, during, and after the event.
- the values of the weight factors may also be adjusted to generate specific temporal weight values for specific points in time.
- the values of the weight factors can be chosen to generate a temporal weight value that is, by way of example, 1 ⁇ 3 rd of the maximum temporal weight value at a point in time that is T 2 ⁇ T 1 (i.e., the duration or length of the event) from T 1 (i.e., the start of the event).
- the information dissemination system calculates a temporal weight of a search result using the following equations:
- TW BEWF* TimeWeight(2.0*( T 1 ⁇ TN )/( TD*TF )) (3)
- TW DEWF* TimeWeight(1.0*((( T 1 +T 2)/2)) ⁇ TN /( TD*TF )) (4)
- TW AEWF* TimeWeight(4.0*( TN ⁇ T 2)/( TD*TF )) (5)
- TimeWeight is the following function:
- TimeWeight( x ) exp( ⁇ 1.0 *abs ( x )) (6)
- the information dissemination system may vary the values that are assigned to the weight factors depending on factors such as the duration of the event, the importance of the event, etc.
- the information dissemination system allows for the creation of event classes and disseminates information related to topics extracted from the event classes.
- Event classes can generally be thought of as scripts that specify topics and actions for obtaining information related to the topics.
- An event class can be written in any of a variety of well-known data description languages such as Extensible Markup Language (XML), etc.
- Event classes are executed to obtain information related to the particular event class.
- the information dissemination system may provide a UI through which an authorized user, such as a system administrator, can create an event class.
- the user can use the UI to specify an event class recognizer, a set of special topics, and, for each specified special topic, a set of slots, a slot recognizer for each slot in the set of slots, and a set of actions.
- the event class recognizer allows for the identification of the corresponding event class.
- the event class recognizer may be specified in the form of heuristics, regular expressions, patterns, or other suitable rules that may be executed to identify its corresponding event class.
- the set of special topics define the relevant ideas for its corresponding event class. For example, a special topic may be created for a travel advisory, a weather advisory, a meeting, a dining experience at a restaurant, and various other events or concepts of events.
- Each slot is a parameter (i.e., a placeholder) that describes an attribute of its corresponding special topic.
- a travel advisory special topic the user may specify slots for an originating city (“FromCity”), a destination city (“ToCity”), one or more connecting cities, an airline carrier, a flight number, a date, and the like.
- ToCity originating city
- ToCity destination city
- connecting cities an airline carrier
- flight number a date
- the like the user may specify slots for a subject of the meeting, invited attendees in the meting, represented companies, and the like.
- the slot recognizer allows for the recognition or identification of a value or “slot filler” for its corresponding slot.
- the slot recognizer may be specified in the form of heuristics, regular expressions, information extraction routines, or other suitable rules that may be executed to identify the appropriate slot filler for the corresponding slot.
- a slot recognizer for a FromCity slot in a travel advisory special topic may specify a restriction or limitation that the FromCity slot can only be filled with a name of a city recognized from a standard or specified list of city names.
- the information dissemination system can then execute the slot recognizer on a set of relevant calendar items, emails, and other information sources of the user to determine the value for the slot limited to any specified restriction.
- the specified actions are invoked to obtain the information corresponding to its special topic.
- the action may be specified as a parameterized URL (i.e., a URL having one or more parameters) that is to be instantiated and invoked.
- a parameterized URL i.e., a URL having one or more parameters
- the information dissemination system first determines a value for each of the parameters included in the URL, and then substitutes the parameters with its corresponding values.
- the action may also be specified as a URL.
- the user may specify rules for digesting and extracting relevant information from the information that is obtained by performing the specified actions.
- the information dissemination system To disseminate the information related to an event class, the information dissemination system first identifies a user event that is of interest to a user. The information dissemination system may identify the user event from the user's profile. The information dissemination system then executes the defined event class recognizers to determine if the user event “fits into” one of the created event classes. Executing the defined event class recognizers may not result in the identification or determination of an event class for the user event. The information dissemination system may allow the user to specify an event class for the user event. Upon determining the event class to execute, the information dissemination system identifies the special topics that have been defined for the event class and the slots that have been defined for each special topic.
- the information dissemination system then identifies the slot recognizers that have been defined for each of the slots and executes each slot recognizer against the various sources of user information and interests, such as the user's emails, appointments, calendar items, etc., to identify the slot filler for the corresponding slot.
- the information dissemination system also instantiates any parameterized actions by identifying the slot filler values for and substituting the values in the appropriate places.
- the information dissemination system then performs the actions defined for the special topic, obtains the action results, and renders the action results to the user. For example, the information dissemination system may instantiate and invoke the defined URLs, and display the results from invoking the URLs.
- the information dissemination system may digest the action results to identify the relevant information and render the relevant information to the user.
- the information dissemination system may rank the action results and renders the ranked action results to the user.
- FIG. 1 is a block diagram that illustrates the processing of the information dissemination system, according to some embodiments.
- the information dissemination system identifies the interests of a user, garners information that is based on the identified user interests, ranks the garnered information, and renders the ranked information to the user. For each user, the information dissemination system first identifies the user's interests and determines the overall and event-specific topics. The user's interests may be identified using information derived from the user's calendar, emails, related documents, etc., and the topics may be determined using various clustering techniques. The information dissemination system then automatically creates a profile for the user.
- the information dissemination system may include in the profile information such as the sources of information to search to obtain the information to render to the user, the events which were analyzed and the topics that were generated, a language model that is augmented with user experience with the information dissemination system, as well as other information regarding the user.
- the information dissemination system uses the information sources and topics specified in the user's profile to periodically searches for and collates the search results containing information that is potentially relevant to the topics.
- the information dissemination system then ranks the search results based on various factors, such as, topic weights, rankings provided by the information sources, the proximity to events, and user feedback information.
- the information dissemination system renders the ranked search results to the user.
- the information dissemination system then obtains user feedback information by monitoring the user's interaction with the rendered search results and the information dissemination itself.
- the information dissemination system then incorporates the feedback information into the user's profile.
- the information dissemination system is able to incorporate the feedback into subsequent searches and the ranking of the search results, thus becoming increasingly more responsive to the user's information needs.
- FIG. 2 is a block diagram that illustrates selected components of the information dissemination system, according to some embodiments.
- the information dissemination system may include both client components that execute on user computing devices, and server components that execute on a server.
- client devices 202 , a server 204 , and information sources 206 are each coupled to a communications link 208 .
- the client devices correspond to computing devices used by users to interact with the information dissemination system.
- Each client device comprises a profiler component 210 .
- the profiler component is invoked to identify the interests of the user and to automatically create a profile for the user.
- the profiler component may then be periodically invoked to update the information regarding the user's interests maintained in the user's profile.
- the profiler component is also invoked to incrementally generate topics that may be of interest to the user.
- the server comprises a profile retriever component 212 , a profile store 214 , an information garnerer component 216 , a ranker component 218 , a renderer component 220 , and a feedback component 222 .
- the profile retriever component is invoked to obtain the user profiles from the client devices.
- the profile store contains the user profiles.
- the information garnerer component is invoked at periodic intervals to garner information from appropriate information sources according to information contained in the user profiles.
- the information garnerer component typically searches the information sources using the queries specified in the user profiles.
- the ranker component is invoked to rank the garnered information (e.g., search results).
- the renderer component is invoked to render ranked search results to appropriate users.
- the feedback component is invoked to obtain user feedback information and to augment the information in the user profiles with the user feedback information.
- the client devices include components of the information dissemination system to allow users interact with the information dissemination system to, for example, view and modify the contents of user profiles, receive notifications and alerts regarding the availability of information, and to display and view the rendered information.
- the server also includes components of the information dissemination system to define event classes and to execute the event classes.
- the computing device on which the information dissemination system is implemented may include a central processing unit, memory, input devices (e.g., keyboard and pointing devices), output devices (e.g., display devices), and storage devices (e.g., disk drives).
- the memory and storage devices are computer-readable media that may contain instructions that implement the information dissemination system.
- the data structures and message structures may be stored or transmitted via a data transmission medium, such as a signal on a communications link.
- Various communication links may be used, such as the Internet, a local area network, a wide area network, a point-to-point dial-up connection, a cell phone network, and so on.
- Embodiments of the information dissemination system may be implemented in various operating environments that include personal computers, server computers, hand-held or laptop devices, multiprocessor systems, microprocessor-based systems, programmable consumer electronics, digital cameras, network PCs, minicomputers, mainframe computers, network devices, distributed computing environments that include any of the above systems or devices, and so on.
- the computer systems may be cell phones, personal digital assistants, smart phones, personal computers, programmable consumer electronics, digital cameras, and so on.
- the information dissemination system may be described in the general context of computer-executable instructions, such as program modules, executed by one or more computers or other devices.
- program modules include routines, programs, objects, components, data structures, and so on that perform particular tasks or implement particular abstract data types.
- functionality of the program modules may be combined or distributed as desired in various embodiments.
- FIG. 3 is a flow diagram that illustrates the processing of the profiler component, according to some embodiments.
- the profiler component periodically identifies the interests of a user to create or update the user's profile with the information regarding the user's interests.
- the profiler component extracts the terms from messages and events contained in selected folders, such as the user's email folders.
- the profiler component generates clusters of similar terms.
- the profiler component associates the clusters with the events.
- the profiler component identifies the user's events, such as the user's appointments.
- the profiler component selects the next identified event.
- the profiler component completes, else the profiler component continues at block 314 .
- the profiler component identifies the clusters that are related to the event.
- the profiler component selects the next identified cluster.
- decision block 318 if all of the identified clusters have already been selected, then the profiler component loops to block 310 to select the next identified event, else the profiler component continues at block 320 .
- the profiler component identifies the key terms in the identified cluster.
- the profiler component generates a query from the key terms.
- the profiler component In block 324 , the profiler component generates a topic for the event and, in block 326 , includes the information regarding the event and the generated query in the topic. In block 328 , the profiler component places the topic in the user's profile, and then loops to block 310 to select the next identified event. If a profile does not exist for the user, then the profiler component creates a profile for the user.
- FIG. 4 is a flow diagram that illustrates the processing of the profiler component to identify key terms, according to some embodiments.
- the profiler component can identify the key terms from the terms that are contained in an event description or an email message.
- the profiler component identifies the key terms based on an overall weight assigned to each term.
- the profiler component selects the next term.
- the profiler component can select the term using any of a variety of well-known tools such as a lexical analysis system.
- decision block 404 if all of the terms have already been selected, then the profiler component continues at block 414 , else the profiler component continues at decision block 406 .
- the profiler component determines a type weight for the term.
- the profiler component calculates a relevance weight for the term.
- the profiler component calculates an overall weight based on the type weight and the relevance weight and assigns the overall weight to the term, and then loops to block 402 to select the next term.
- the profiler component identifies the key terms based on the overall weights assigned to the terms, and completes.
- FIG. 5 is a flow diagram that illustrates the processing of the profiler component to generate a query, according to some embodiments.
- the profiler component generates a query from the key terms.
- the profiler component ranks the key terms based on their overall weights.
- the profiler component identifies the six highest ranking key terms.
- the profiler component generates a query from the identified six highest ranking key terms using Equation 1, and completes. In other embodiments, the profiler component may generate a query using a different number of the highest ranking key terms.
- FIG. 6 is a flow diagram that illustrates the processing of the profiler component to incrementally generate a topic, according to some embodiments.
- the profiler component receives an indication that an event is of interest to a user. For example, the user may specify that the event is newsworthy and submit a request to receive information related to the specified event.
- the profiler component identifies a user interest, such as a relevant message or an event.
- the profiler component extracts the terms associated with the identified user interest.
- the profiler component identifies the key terms from the extracted terms.
- the profiler component generates a query from the key terms.
- the profiler component places the query in the user's profile, and completes.
- FIG. 7 is a flow diagram that illustrates the processing of the information garnerer component, according to some embodiments.
- the information garnerer component periodically garners information for rendering to a user based on the information contained in the user's profile.
- the garnerer component selects the next query from the user's profile.
- decision block 704 if all of the queries have already been selected, then the garnerer component continues at block 708 , else the garnerer component continues at block 706 .
- the garnerer component executes the query and obtains the search results.
- the garner component may execute the query by searching the information sources specified in the user's profile using the query.
- the garnerer component loops to block 702 to select the next query.
- the garnerer component ranks the search results which were obtained by executing the queries.
- the garnerer component renders the ranked search results to the user, and completes.
- FIG. 8 is a flow diagram that illustrates the processing of the ranker component, according to some embodiments.
- the ranker component is passed the search results (the information items resulting from a search of an information source) and ranks the search results based on a total weight assigned to each search result.
- the ranker component selects the next search result.
- decision block 804 if all of the search results have already been selected, then the ranker component continues at block 814 , else the ranker component continues at block 806 .
- the ranker component calculates a static weight for the search result.
- the ranker component calculates a temporal weight for the search result.
- the ranker component calculates an adaptive weight for the search result.
- the ranker component calculates a total weight based on the static weight, temporal weight, and the adaptive weight, assigns the total weight to the search result, and then loops to block 802 to select the next search result.
- the ranker component ranks the search results based on the total weights assigned to the search results, and completes.
- FIG. 9 is a graphical depiction of weight values as a function of proximity to an event, according to some embodiments.
- the graph illustrates the temporal weight value, which is the importance of an item of information (i.e., search result) to a user, relative to time.
- the relative importance of the item of information varies differently depending on the current time.
- T 1 is the scheduled start time of the event related to the item of information
- T 2 is the scheduled end time of the related event.
- the importance of the item of information to the user rises, for example, exponentially as depicted in FIG. 9 , but with a shallow upward curve as the current time approaches closer to T 1 .
- the importance of the item of information to the user remains relatively flat.
- the importance of the item of information to the user quickly drops, for example, exponentially as depicted in FIG. 9 , so that the importance diminishes to a value of zero in a very short amount of time.
- the importance of the item of information to the user during the event can rise to a peak at a point in time during the event, for example at time (T 1 +T 2 )/2, and decrease thereafter to the scheduled end of the event.
- FIG. 11 is a flow diagram that illustrates the processing of the ranker component to calculate a temporal weight of a search result, according to some embodiments.
- the ranker component identifies the event that corresponds to the search result.
- the ranker component determines the current time (i.e., the time at the present).
- the ranker component determines the time the identified event is scheduled to start.
- the ranker component determines the time the identified event is scheduled to end.
- decision block 1110 if the current time is before the time the event is scheduled to start, then the ranker component continues at block 1112 , else the ranker component continues at decision block 1114 .
- the ranker component calculates a temporal weight for the search result using Equation 3, and completes.
- decision block 1114 if the current time is between the time the event is scheduled to start and the time the event is scheduled to end, then the ranker component continues at block 1116 , else the ranker component continues at block 1118 .
- the ranker component calculates a temporal weight for the search result using Equation 4, and completes.
- the ranker component calculates a temporal weight for the search result using Equation 5, and completes.
- FIG. 12 is a flow diagram that illustrates the rendering information based on an event class corresponding to a user event, according to some embodiments.
- the information dissemination system creates event classes and disseminates information related to topics extracted from the event classes.
- a component of the information dissemination system creates event classes.
- the component identifies a user event to process.
- the component determines an event class that corresponds to the identified user event.
- the component processes the identified event class, and completes.
- FIG. 13 is a flow diagram that illustrates the creating of an event class, according to some embodiments.
- a component of the information dissemination system may create an event class in response to a user request to create an event class and based upon input provided by the user.
- the component creates an event class recognizer for the event class.
- the component creates a set of special topics for the event class.
- the component selects the next special topic.
- decision block 1308 if all of the special topics have already been selected, then the component completes, else the component continues at block 1310 .
- the component creates a set of slots for the special topic.
- the component creates a slot recognizer for each of the created slots.
- the component creates a set of actions for the special topic.
- the component creates a set of rules for digesting the action results to extract the relevant information from the action results, and then loops to block 1306 to select the next special topic.
- FIG. 14 is an example data listing that illustrates an event class, according to some embodiments.
- the listing illustrates a plurality of special topics contained in the event class, including a “Travel advisory topic” section 1402 for a travel advisory special topic.
- the travel advisory topic section contains a “Slots” section 1404 , a “Recognizers” section 1406 , and an “Action” section 1408 .
- the slots section specifies the slots that have been defined for the special topic. For example, the slots “ToCity” and “ToCountry” have been defined for the travel advisory topic.
- the recognizers section specifies the restrictions on the values of the slots and how to determine the values for the slots.
- the action section specifies the actions that are to be taken for the special topic.
- the actions may be specified in the form of parameterized URLs to be displayed or executed.
- the listing in FIG. 14 also illustrates other special topics which have not been illustrated to the same level of detail as the travel advisory special topic. Although not depicted, each of these special topics may also contain sections similar to those contained in the travel advisory special topic.
- FIG. 15 is a flow diagram that illustrates processing an event class, according to some embodiments.
- a component of the information dissemination system may process an event class upon identifying a user event to process.
- the component identifies the special topics that have been defined (i.e., created) for the event class.
- the component selects the next defined special topic.
- decision block 1506 if all of the defined special topics have already been selected, then the component completes, else the component continues at block 1508 .
- the component identifies the slots that have been defined for the special topic.
- the component selects the next defined slot.
- the component uses the slot recognizer for the slot to identify the slot filler (and a value) for the slot, and then loops to block 1510 to select the next defined slot. In some embodiments, in instances where the component is unable to identify the slot filler, the component may request the user to provide the slot filler.
- the component identifies the actions that have been defined for the special topic.
- the component performs the actions.
- the component digests the action results, for example, to identify the relevant information.
- the component renders the digested action results to the user, and then loops to block 1504 to select the next defined special topic.
Abstract
An information dissemination system ranks the search results based on a temporal weight assigned to each search result. The temporal weight is an indication of the importance to a user that varies with time. For each search result, the information dissemination system calculates a temporal weight that is based on the temporal proximity of the event that is related to the search result. The temporal weight may be used to re-rank the search results.
Description
- This application is related to U.S. patent application Ser. No. ______, entitled “Presenting Information Related to Topics Extracted from Event Classes,” which was filed on even date herewith, and is identified by attorney docket number 418268343US, and U.S. patent application Ser. No. ______, entitled “Adaptive Dissemination of Personalized and Contextually Relevant Information,” which was filed on even date herewith, and is identified by attorney docket number 418268343US2, the disclosures of which are incorporated by reference herein in their entireties.
- The development of information systems, such as the Internet, and various on-line services for accessing the information systems, has led to the availability of increasing amounts of information. As computers become increasingly powerful and versatile, users are increasingly employing their computers for a broad variety of tasks. Accompanying the increasing use and versatility of computers is an increasing desire on the part of users to increasingly rely on their computing devices to perform their daily activities. For example, anyone with access to a suitable Internet connection may go “online” and navigate to the information pages (i.e., the web pages) to gather information that is relevant to the user's current activity.
- Many search engine services, such as Google and Yahoo!, provide for searching for information that is accessible via the Internet. These search engine services allow users to search for display pages, such as web pages, that may be of interest to users. After a user submits a search request (i.e., a query) that includes search terms, the search engine service identifies web pages that may be related to those search terms. To quickly identify related web pages, the search engine services may maintain a mapping of keywords to web pages. This mapping may be generated by “crawling” the web (i.e., the World Wide Web) to identify the keywords of each web page. To crawl the web, a search engine service may use a list of root web pages to identify all web pages that are accessible through those root web pages. The keywords of any particular web page can be identified using various well-known information retrieval techniques, such as identifying the words of a headline, the words supplied in the metadata of the web page, the words that are highlighted, and so on. The search engine service may generate a relevance score to indicate how relevant the information of the web page may be to the search request based on the closeness of each match, web page importance or popularity (e.g., Google's PageRank), and so on. The search engine service then displays to the user links to those web pages in an order that is based on a ranking determined by their relevance.
- Unfortunately, users of the information systems encounter an information overload problem. For example, the search engine services often provide its user irrelevant search results, thus forcing the user to sift through a long list of web pages in order to find the relevant web pages. Compounding this problem is the constantly changing and expanding volume of information, and the difficulty users encounter in keeping abreast of the constantly changing information to determine the information that is credible and relevant to the user.
- A method and system for ranking search results based on the temporal proximity to the event to which the search results relate is provided. An information dissemination system ranks the search results based on a temporal weight assigned to each search result. The temporal weight is an indication of the importance to a user that varies with time. For each search result, the information dissemination system calculates a temporal weight that is based on the temporal proximity of the event that is related to the search result. The temporal weight may be used to re-rank the search results.
- This Summary is provided to introduce a selection of concepts in a simplified form that are further described below in the Detailed Description. This Summary is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used as an aid in determining the scope of the claimed subject matter.
-
FIG. 1 is a block diagram that illustrates the processing of the information dissemination system, according to some embodiments. -
FIG. 2 is a block diagram that illustrates selected components of the information dissemination system, according to some embodiments. -
FIG. 3 is a flow diagram that illustrates the processing of the profiler component, according to some embodiments. -
FIG. 4 is a flow diagram that illustrates the processing of the profiler component to identify key terms, according to some embodiments. -
FIG. 5 is a flow diagram that illustrates the processing of the profiler component to generate a query, according to some embodiments. -
FIG. 6 is a flow diagram that illustrates the processing of the profiler component to incrementally generate a topic, according to some embodiments. -
FIG. 7 is a flow diagram that illustrates the processing of the information garnerer component, according to some embodiments. -
FIG. 8 is a flow diagram that illustrates the processing of the ranker component, according to some embodiments. -
FIG. 9 is a graphical depiction of weight values as a function of proximity to an event, according to some embodiments. -
FIG. 10 is a graphical depiction of weight values as a function of proximity to an event, according to some embodiments. -
FIG. 11 is a flow diagram that illustrates the processing of the ranker component to calculate a temporal weight of a search result, according to some embodiments. -
FIG. 12 is a flow diagram that illustrates the rendering information based on an event class corresponding to a user event, according to some embodiments. -
FIG. 13 is a flow diagram that illustrates the creating of an event class, according to some embodiments. -
FIG. 14 is an example data listing that illustrates an event class, according to some embodiments. -
FIG. 15 is a flow diagram that illustrates processing an event class, according to some embodiments. - A method and system for adaptively disseminating personalized and contextually relevant information is provided. In some embodiments, an information dissemination system identifies the interests of a user and adaptively disseminates to the user the information that is relevant to the user's interests. The user's interests may be identified from information derived from various sources, such as, by way of example, the user's calendar entries, email messages, World Wide Web (web) usage information, documents, enterprise resource planning (ERP) data, and the like. The information dissemination system then creates a profile for the user, and includes in the profile the details about the user's interests. For example, the information dissemination system may include in the user's profile information regarding: the information sources to query for information to render to the user; the user events (e.g., meetings, appointments, etc.) which were analyzed to determine the user's interests; the topics derived from the analyzed events; the queries generated for the analyzed events; a language model of terms that are relevant to the user; the user's preferences for receiving the relevant information; and other information regarding the user. Generally, the topic can be defined as any cohesive concept of the event. For example, for a conference that the user is scheduled to attend, the topics may include the subject of the conference, the people presenting in the conference, the people attending the conference, the location of the conference, the organizations represented at the conference, travel information for the user (e.g., origination city, destination city, travel mode, flight numbers, hotel reservations, rental car reservations, etc.), and weather information. The terms that are relevant to the user may be derived from the various sources used to identify the user's interests. The information dissemination system periodically executes the queries in the user's profile against the appropriate information sources and obtains the search results. As used herein, the term “search result” or any variant thereof generally refers to results which are filtered using a query. The information dissemination system then ranks the search results according to various factors, and renders the ranked search results to the user. The information dissemination system obtains feedback from the user regarding the rendered search results and incorporates the feedback into the user's profile. The information dissemination system may provide a user interface (UI) through which the user can view and/or modify the details of the user's interests maintained in the user's profile. For example, a user may use the UI to specify the level of importance and/or relevance of the events, topics, terms, information sources, and other information included the user's profile. The information dissemination system may also monitor the user's interaction with the rendered search results and make appropriate modifications to the details of the user's interests maintained in the user's profile. The information dissemination system may also periodically update the user's profile with more recent or updated details about the user's interests. In this manner, the information dissemination system, over time, becomes increasingly responsive to the user's information needs.
- In some embodiments, the information dissemination system identifies key terms from the terms that are relevant to a user and generates the queries from the key terms. The information dissemination system can identify the key terms from the terms extracted from the user's messages and events contained in selected folders. By way of example, the information dissemination system may scan the user's recent email folders (e.g., inbox folder and sent items folder) and extract the terms from these email folders. The information dissemination system can then identify the emails in the other folders that are related to the extracted terms, and extract the terms from these related emails. Using the extracted terms, the information dissemination system creates clusters of terms that are similar. Each cluster is associated with a set of terms. The information dissemination system then identifies events, such as the user's appointments by scanning the entries in a calendaring application, and, for each event, identifies the clusters associated with the event. The information dissemination system then creates one or more topics for each event. Each topic may be considered a “container” that contains information that is relevant to its associated event, such as an event name, an event duration, etc. The information dissemination system identifies the key terms in the identified clusters and generates a query from the key terms. The information dissemination system then places the generated query into the topic that is associated with the event, and includes the topic in the user's profile.
- The information dissemination system may use various techniques to generate the clusters of terms. In some embodiments, the information dissemination system uses a K-means algorithm. The information dissemination system provides to the algorithm a number K that specifies the desired number of clusters. In the first pass, the algorithm takes the first K terms as the centroid of a unique cluster. The centroid is the mean position of the cluster (e.g., the mean position of the terms in the cluster). The algorithm then assigns each of the remaining terms to the cluster with the closest centroid to the term (e.g., the cluster where the difference between the centroid and the term is minimized). In the subsequent passes, the algorithm recomputes the cluster centroids based the previous pass and then reassigns each term to the cluster with the centroid that is closest to the term. The algorithm may be performed for a fixed number of passes (e.g., three) or until the clustering converges on a solution. The information dissemination system may use any of a variety of well-known algorithms, including maximum likelihood estimation, spectral clustering, and so on, to generate the clusters of terms.
- In some embodiments, the information dissemination system identifies the key terms in a cluster based on an overall weight assigned to each term. The overall weight of a term may be derived from a combination of a type weight of the term and a relevance weight of the term. The information dissemination system first filters the terms to reduce the number of terms that may be identified as key terms. For example, the information dissemination system may filter out the terms that are Uniform Resource Locators (URLs), and terms that are made up of numbers in order to reduce the number of candidate key terms. For each of the remaining terms, the information dissemination system can calculate a type weight and a relevance weight. The type weight is a weight that is assigned to a term based on the type of the term. For example, the information dissemination system may assign a larger type weight to a term that is a name, such as a name of a person, than to a term that is comprised of multiple words. A term that is comprised of multiple words may in turn be assigned a larger type weight than a term that is a single word. The information dissemination system may also assign different type weights to a term based on whether the term appears as part of the subject text, body text, contacts text, or other parts of the event. The relevance weight of a term is a measure of relevance of the term to the user and, in particular, to the user's interests. One common technique for determining the relevance of the term to the user is based on term frequency and inverse document frequency (TF*IDF). Term frequency refers to the number of occurrences of a term within a document, and the inverse document frequency refers to the inverse of the number of documents that contain that term. In this instance, the documents generally refer to the various information sources used to identify the user's interests. The information dissemination system may identify the terms that are over represented and the terms that are underrepresented as terms that are of unlikely relevance to the user. To each of these terms, the information dissemination system assigns a low relevance weight. To each of the remaining terms, the information dissemination system can assign a relevance weight based upon the term's TF*IDF metric. The information dissemination system may use any of a variety of other well-known techniques for determining the relevance of a term. The information dissemination system may then derive an overall weight for a term based on the term's type weight and relevance weight. For example, the overall weight of a term may be derived by multiplying the type weight and the relevance weight of the term. In some embodiments, the overall weight of a term may be derived from the term's relevance weight. The information dissemination system may then identify as key terms the terms whose overall weights are above a predetermined threshold. The information dissemination system may maintain the terms and their assigned weights in the user's profile, for example, as part of the language model.
- In some embodiments, the information dissemination system generates a query from the key terms by using the highest six ranking key terms. The information dissemination system first ranks the key terms based on the overall weights assigned to the key terms. The information dissemination system then identifies the highest six ranking key terms and generates the query by the following equation:
-
Query=A AND B AND (C OR OR OR F) (1) - where A represents the highest ranking key term, represents the second highest ranking key term, C represents the third highest ranking key term, D represents the fourth highest ranking key term, E represents the fifth highest ranking key term, and F represents the sixth highest ranking key term. In instances where there are less than six key terms, the information dissemination system may generate the query using the number of available key terms. For example, if there are only four key terms, the information dissemination system can generate the query by omitting the fifth and sixth highest ranking key terms in equation (1) above. Examples of other suitable queries where there are less than six key terms include: Query=A; Query=A AND B; Query=A AND B AND C; Query=A AND B AND (C OR D); etc.
- In some embodiments, the information dissemination system incrementally generates topics that may be of interest to a user. For example, the user may provide an indication that some event is newsworthy (i.e., of interest to the user). Upon receiving the indication, the information dissemination system identifies a user interest. For example, the information dissemination system can identify a topic from the specified event as being of interest to the user. Any one of the other topics that are related to the specified event may also be identified as the user interest. The information dissemination system then extracts the terms that are associated with the topic and, from these terms, identifies the key terms. The information dissemination system can identify the key terms based on the overall weights assigned to the terms, as described above. The information dissemination system then generates a query from the key terms and includes the query in the user's profile for execution. The information dissemination system may also maintain the extracted terms and their assigned weights in the user's profile, for example, as part of the language model.
- The information dissemination system garners information for a user at periodic intervals. At a periodic interval, the information dissemination system executes the queries against the information sources specified in the user's profile to obtain search results. The information sources may include MSN Search, MSN News Search, and other information sources which index blogs, news feeds, web pages, document sources, webcasts, videocasts, audiocasts, etc., and which makes the search results available via an interface. The information dissemination system then generates a ranked list of information items by ranking the items of information included in the search results. The information dissemination system then renders the ranked search results to the user. The information dissemination system initializes the periodic intervals to default values, which may be modified by the user. In some embodiments, the information dissemination system garners information related to events that are within a time range, such as a date range. For example, the information dissemination may initialize the date range to one week (e.g., a period of seven days starting from the day before the current day (today)). In this instance, the information dissemination system garners information by executing the queries generated for the topics corresponding to the events that are within the date range. In some embodiments, the information dissemination system allows the user to initiate the garnering of information. For example, the information dissemination system may provide a UI through which the user can specify a date range of interest and request the garnering of information.
- In some embodiments, the information dissemination system ranks the search results based on a total weight assigned to each search result. The total weight of a search result may be derived from a combination of a static weight of the search result, an adaptive weight of the search result, and a temporal weight of the search result. The static weight may be derived from various weighting factors such as the importance of the topic/event to the user, the importance of the information source, the ranking provided by the information source, and the relevance of the terms to the user's interests. For example, a static weight for a search result may be a combination of the importance of the topic that corresponds to the search result, the importance of the information source that produced the search result, the rank provided by the information source that produced the search result, and the relevance of the terms in the search result to the user's interests. The relevance of the terms may be determined using the language model included in the user's profile. For example, a cosine similarity measure can be employed to measure the similarity of the terms in the search result to the terms in the language model. The adaptive weight is a weight that is assigned to a search result based on user preferences and/or feedback. For example, the user may specify additional information sources from which to retrieve information. The user may indicate a preference order for the information sources. The user may also indicate a preference for a certain type of search result over another type of search result. The information dissemination system may also discern the user's preferences by monitoring the user's interactions with the information dissemination system and, in particular, the information rendered by the information dissemination system. The information dissemination system may maintain information regarding the user's preferences and interactions in the user's profile. The temporal weight is a weight that is assigned to a search result based on the proximity of the event that corresponds to the search result. Temporal weighting is further discussed below. The information dissemination system may then derive a total weight for each search result based on the search result's static weight, adaptive weight, and temporal weight. For example, the total weight of a search result may be derived by adding the static weight, adaptive weight, and temporal weight of the search result. The information dissemination system then ranks the search results according to their total weights and renders the ranked list of search results. In some embodiments, the information dissemination system renders only a predetermined number of search results. For example, the information dissemination system may only render the highest twenty ranking search results.
- In some embodiments, the information dissemination system may alert users regarding highly relevant search results. The information dissemination system may alert a user by providing an indication on the user's client computing device. For example, the information dissemination system may cause a particular icon or indicator to appear on a UI on the user's client computing device. The information dissemination system may also vary the color of a displayed icon or indicator to indicate the presence of search results of varying relevance. For a user using a mobile client computing device, such as smart phone, the information dissemination system may send the user a text message informing the user of the presence of relevant search results.
- Temporal ranking is based on the notion that the importance of information varies with time. By way of example, a user may have two events scheduled in the user's calendar: a first event regarding a meeting next week in New York, and a second event regarding the flight to New York for this meeting. At some point in time prior to the meeting, the importance of the meeting event will be high (i.e., of interest to the user), and any information items (e.g., news) related to the meeting event will also be high. As the time approaches the flight event, the flight event also increases in importance, and information items related to the flight event, such as flight delay information, also increases in importance. Once the user takes the flight and lands in New York, the flight event will have vastly diminished in importance to the user, and related information items may no longer be of interest to the user. As can be seen from this example, the events have importance to the user, and information items related to the events also have importance to the user, but this importance varies according to time and, more specifically, the proximity to the events.
- In some embodiments, the information dissemination system ranks the search results based on a temporal weight assigned to each search result. The temporal weight is an indication of importance of a search result to a user that varies with time. For each search result, the information dissemination system calculates a temporal weight that is based on the temporal proximity of the event that is related to the search result. The value of the temporal weight of a search result may increase (e.g., rise) exponentially in the time period before the start of the related event. During the related event, the value of the temporal weight may remain relatively constant (e.g., flat), or may rise to a peak at a point in time during the related event and decrease thereafter. In the time period after the end of the related event, the value of the temporal weight may decrease (e.g., drop) exponentially to zero in a very short amount of time. The value of the temporal weight may be calculated using an exponential function of the following form:
-
f(t)=k*exp(−m*rti) (2) - where rti represents a time interval (e.g., time from the current time to the start of the event) compared to the length of the event (e.g., T2−T1, where T2 is the end time of the event and T1 is the start time of the event), and k and m are weight factors whose values are chosen to vary the shape of the function as desired. The values chosen for the weight factors k and m may be different before, during, and after the event. The values of the weight factors may also be adjusted to generate specific temporal weight values for specific points in time. For example, the values of the weight factors can be chosen to generate a temporal weight value that is, by way of example, ⅓rd of the maximum temporal weight value at a point in time that is T2−T1 (i.e., the duration or length of the event) from T1 (i.e., the start of the event).
- In some embodiments, the information dissemination system calculates a temporal weight of a search result using the following equations:
- if the current time is before the start of the event related to the search result,
-
TW=BEWF*TimeWeight(2.0*(T1−TN)/(TD*TF)) (3) - if the current time is during the event related to the search result,
-
TW=DEWF*TimeWeight(1.0*(((T1+T2)/2))−TN/(TD*TF)) (4) - if the current time is after the end of the event related to the search result,
-
TW=AEWF*TimeWeight(4.0*(TN−T2)/(TD*TF)) (5) - where TN is the current time, T1 is the time the event starts, T2 is the time the event ends, TD is the length in time of the event, BEWF is a weight factor before the event, DEWF is a weight factor during the event, AEWF is a weight factor after the event, TF is a weight factor, and TimeWeight is the following function:
-
TimeWeight(x)=exp(−1.0*abs(x)) (6) - where exp represents the exponential function “e”, and abs represents the absolute value function. The values assigned to the weight factors BEWF, DEWF, AEWF, and TF control the rate of the rise and the rate of the decline of the temporal weight. In some embodiments, the information dissemination system may vary the values that are assigned to the weight factors depending on factors such as the duration of the event, the importance of the event, etc.
- In some embodiments, the information dissemination system allows for the creation of event classes and disseminates information related to topics extracted from the event classes. Event classes can generally be thought of as scripts that specify topics and actions for obtaining information related to the topics. An event class can be written in any of a variety of well-known data description languages such as Extensible Markup Language (XML), etc. Event classes are executed to obtain information related to the particular event class. In some embodiments, the information dissemination system may provide a UI through which an authorized user, such as a system administrator, can create an event class. To create an event class, the user can use the UI to specify an event class recognizer, a set of special topics, and, for each specified special topic, a set of slots, a slot recognizer for each slot in the set of slots, and a set of actions. The event class recognizer allows for the identification of the corresponding event class. The event class recognizer may be specified in the form of heuristics, regular expressions, patterns, or other suitable rules that may be executed to identify its corresponding event class. The set of special topics define the relevant ideas for its corresponding event class. For example, a special topic may be created for a travel advisory, a weather advisory, a meeting, a dining experience at a restaurant, and various other events or concepts of events. Each slot is a parameter (i.e., a placeholder) that describes an attribute of its corresponding special topic. For example, to create a travel advisory special topic, the user may specify slots for an originating city (“FromCity”), a destination city (“ToCity”), one or more connecting cities, an airline carrier, a flight number, a date, and the like. To create a meeting special topic, the user may specify slots for a subject of the meeting, invited attendees in the meting, represented companies, and the like. The slot recognizer allows for the recognition or identification of a value or “slot filler” for its corresponding slot. The slot recognizer may be specified in the form of heuristics, regular expressions, information extraction routines, or other suitable rules that may be executed to identify the appropriate slot filler for the corresponding slot. For example, a slot recognizer for a FromCity slot in a travel advisory special topic may specify a restriction or limitation that the FromCity slot can only be filled with a name of a city recognized from a standard or specified list of city names. The information dissemination system can then execute the slot recognizer on a set of relevant calendar items, emails, and other information sources of the user to determine the value for the slot limited to any specified restriction. The specified actions are invoked to obtain the information corresponding to its special topic. The action may be specified as a parameterized URL (i.e., a URL having one or more parameters) that is to be instantiated and invoked. To instantiate a parameterized URL, the information dissemination system first determines a value for each of the parameters included in the URL, and then substitutes the parameters with its corresponding values. The action may also be specified as a URL. In some embodiments, the user may specify rules for digesting and extracting relevant information from the information that is obtained by performing the specified actions.
- To disseminate the information related to an event class, the information dissemination system first identifies a user event that is of interest to a user. The information dissemination system may identify the user event from the user's profile. The information dissemination system then executes the defined event class recognizers to determine if the user event “fits into” one of the created event classes. Executing the defined event class recognizers may not result in the identification or determination of an event class for the user event. The information dissemination system may allow the user to specify an event class for the user event. Upon determining the event class to execute, the information dissemination system identifies the special topics that have been defined for the event class and the slots that have been defined for each special topic. The information dissemination system then identifies the slot recognizers that have been defined for each of the slots and executes each slot recognizer against the various sources of user information and interests, such as the user's emails, appointments, calendar items, etc., to identify the slot filler for the corresponding slot. The information dissemination system also instantiates any parameterized actions by identifying the slot filler values for and substituting the values in the appropriate places. The information dissemination system then performs the actions defined for the special topic, obtains the action results, and renders the action results to the user. For example, the information dissemination system may instantiate and invoke the defined URLs, and display the results from invoking the URLs. In some embodiments, the information dissemination system may digest the action results to identify the relevant information and render the relevant information to the user. In some embodiments, the information dissemination system may rank the action results and renders the ranked action results to the user.
-
FIG. 1 is a block diagram that illustrates the processing of the information dissemination system, according to some embodiments. The information dissemination system identifies the interests of a user, garners information that is based on the identified user interests, ranks the garnered information, and renders the ranked information to the user. For each user, the information dissemination system first identifies the user's interests and determines the overall and event-specific topics. The user's interests may be identified using information derived from the user's calendar, emails, related documents, etc., and the topics may be determined using various clustering techniques. The information dissemination system then automatically creates a profile for the user. The information dissemination system may include in the profile information such as the sources of information to search to obtain the information to render to the user, the events which were analyzed and the topics that were generated, a language model that is augmented with user experience with the information dissemination system, as well as other information regarding the user. Using the information sources and topics specified in the user's profile, the information dissemination system periodically searches for and collates the search results containing information that is potentially relevant to the topics. The information dissemination system then ranks the search results based on various factors, such as, topic weights, rankings provided by the information sources, the proximity to events, and user feedback information. The information dissemination system then renders the ranked search results to the user. The information dissemination system then obtains user feedback information by monitoring the user's interaction with the rendered search results and the information dissemination itself. The information dissemination system then incorporates the feedback information into the user's profile. By incorporating the feedback information into the user's profile, the information dissemination system is able to incorporate the feedback into subsequent searches and the ranking of the search results, thus becoming increasingly more responsive to the user's information needs. -
FIG. 2 is a block diagram that illustrates selected components of the information dissemination system, according to some embodiments. The information dissemination system may include both client components that execute on user computing devices, and server components that execute on a server. As depicted,client devices 202, aserver 204, andinformation sources 206 are each coupled to acommunications link 208. The client devices correspond to computing devices used by users to interact with the information dissemination system. Each client device comprises aprofiler component 210. When a user first registers with the information dissemination system, the profiler component is invoked to identify the interests of the user and to automatically create a profile for the user. The profiler component may then be periodically invoked to update the information regarding the user's interests maintained in the user's profile. The profiler component is also invoked to incrementally generate topics that may be of interest to the user. The server comprises aprofile retriever component 212, aprofile store 214, aninformation garnerer component 216, aranker component 218, arenderer component 220, and afeedback component 222. The profile retriever component is invoked to obtain the user profiles from the client devices. The profile store contains the user profiles. The information garnerer component is invoked at periodic intervals to garner information from appropriate information sources according to information contained in the user profiles. The information garnerer component typically searches the information sources using the queries specified in the user profiles. The ranker component is invoked to rank the garnered information (e.g., search results). The renderer component is invoked to render ranked search results to appropriate users. The feedback component is invoked to obtain user feedback information and to augment the information in the user profiles with the user feedback information. Although not shown inFIG. 2 , the client devices include components of the information dissemination system to allow users interact with the information dissemination system to, for example, view and modify the contents of user profiles, receive notifications and alerts regarding the availability of information, and to display and view the rendered information. Similarly, the server also includes components of the information dissemination system to define event classes and to execute the event classes. - The computing device on which the information dissemination system is implemented may include a central processing unit, memory, input devices (e.g., keyboard and pointing devices), output devices (e.g., display devices), and storage devices (e.g., disk drives). The memory and storage devices are computer-readable media that may contain instructions that implement the information dissemination system. In addition, the data structures and message structures may be stored or transmitted via a data transmission medium, such as a signal on a communications link. Various communication links may be used, such as the Internet, a local area network, a wide area network, a point-to-point dial-up connection, a cell phone network, and so on.
- Embodiments of the information dissemination system, including the information dissemination client and the information dissemination server may be implemented in various operating environments that include personal computers, server computers, hand-held or laptop devices, multiprocessor systems, microprocessor-based systems, programmable consumer electronics, digital cameras, network PCs, minicomputers, mainframe computers, network devices, distributed computing environments that include any of the above systems or devices, and so on. The computer systems may be cell phones, personal digital assistants, smart phones, personal computers, programmable consumer electronics, digital cameras, and so on.
- The information dissemination system may be described in the general context of computer-executable instructions, such as program modules, executed by one or more computers or other devices. Generally, program modules include routines, programs, objects, components, data structures, and so on that perform particular tasks or implement particular abstract data types. Typically, the functionality of the program modules may be combined or distributed as desired in various embodiments.
-
FIG. 3 is a flow diagram that illustrates the processing of the profiler component, according to some embodiments. The profiler component periodically identifies the interests of a user to create or update the user's profile with the information regarding the user's interests. Inblock 302, the profiler component extracts the terms from messages and events contained in selected folders, such as the user's email folders. Inblock 304, the profiler component generates clusters of similar terms. Inblock 306, the profiler component associates the clusters with the events. Inblock 308, the profiler component identifies the user's events, such as the user's appointments. Inblock 310, the profiler component selects the next identified event. Indecision block 312, if all of the identified events have already been selected, then the profiler component completes, else the profiler component continues atblock 314. Inblock 314, the profiler component identifies the clusters that are related to the event. Inblock 316, the profiler component selects the next identified cluster. Indecision block 318, if all of the identified clusters have already been selected, then the profiler component loops to block 310 to select the next identified event, else the profiler component continues atblock 320. Inblock 320, the profiler component identifies the key terms in the identified cluster. Inblock 322, the profiler component generates a query from the key terms. Inblock 324, the profiler component generates a topic for the event and, inblock 326, includes the information regarding the event and the generated query in the topic. Inblock 328, the profiler component places the topic in the user's profile, and then loops to block 310 to select the next identified event. If a profile does not exist for the user, then the profiler component creates a profile for the user. - One skilled in the art will appreciate that, for this and other processes and methods disclosed herein, the functions performed in the processes and methods may be implemented in differing order. Furthermore, the outlined steps are only exemplary, and some of the steps may be optional, combined with fewer steps, or expanded into additional steps.
-
FIG. 4 is a flow diagram that illustrates the processing of the profiler component to identify key terms, according to some embodiments. For example, the profiler component can identify the key terms from the terms that are contained in an event description or an email message. The profiler component identifies the key terms based on an overall weight assigned to each term. Inblock 402, the profiler component selects the next term. The profiler component can select the term using any of a variety of well-known tools such as a lexical analysis system. Indecision block 404, if all of the terms have already been selected, then the profiler component continues atblock 414, else the profiler component continues atdecision block 406. Indecision block 406, if the term should be eliminated as a potential key term, then the profiler component loops to block 402 to select the next term, else the profiler component continues atblock 408. Inblock 408, the profiler component calculates a type weight for the term. Inblock 410, the profiler component calculates a relevance weight for the term. Inblock 412, the profiler component calculates an overall weight based on the type weight and the relevance weight and assigns the overall weight to the term, and then loops to block 402 to select the next term. Inblock 414, the profiler component identifies the key terms based on the overall weights assigned to the terms, and completes. -
FIG. 5 is a flow diagram that illustrates the processing of the profiler component to generate a query, according to some embodiments. The profiler component generates a query from the key terms. Inblock 502, the profiler component ranks the key terms based on their overall weights. Inblock 504, the profiler component identifies the six highest ranking key terms. Inblock 506, the profiler component generates a query from the identified six highest ranking keyterms using Equation 1, and completes. In other embodiments, the profiler component may generate a query using a different number of the highest ranking key terms. -
FIG. 6 is a flow diagram that illustrates the processing of the profiler component to incrementally generate a topic, according to some embodiments. The profiler component receives an indication that an event is of interest to a user. For example, the user may specify that the event is newsworthy and submit a request to receive information related to the specified event. Inblock 602, the profiler component identifies a user interest, such as a relevant message or an event. Inblock 604, the profiler component extracts the terms associated with the identified user interest. Inblock 606, the profiler component identifies the key terms from the extracted terms. Inblock 608, the profiler component generates a query from the key terms. Inblock 610, the profiler component places the query in the user's profile, and completes. -
FIG. 7 is a flow diagram that illustrates the processing of the information garnerer component, according to some embodiments. The information garnerer component periodically garners information for rendering to a user based on the information contained in the user's profile. Inblock 702, the garnerer component selects the next query from the user's profile. Indecision block 704, if all of the queries have already been selected, then the garnerer component continues atblock 708, else the garnerer component continues atblock 706. Inblock 706, the garnerer component executes the query and obtains the search results. The garner component may execute the query by searching the information sources specified in the user's profile using the query. Subsequent to obtaining the search results for the query, the garnerer component loops to block 702 to select the next query. Inblock 708, the garnerer component ranks the search results which were obtained by executing the queries. Inblock 710, the garnerer component renders the ranked search results to the user, and completes. -
FIG. 8 is a flow diagram that illustrates the processing of the ranker component, according to some embodiments. The ranker component is passed the search results (the information items resulting from a search of an information source) and ranks the search results based on a total weight assigned to each search result. Inblock 802, the ranker component selects the next search result. Indecision block 804, if all of the search results have already been selected, then the ranker component continues atblock 814, else the ranker component continues atblock 806. Inblock 806, the ranker component calculates a static weight for the search result. Inblock 808, the ranker component calculates a temporal weight for the search result. Inblock 810, the ranker component calculates an adaptive weight for the search result. Inblock 812, the ranker component calculates a total weight based on the static weight, temporal weight, and the adaptive weight, assigns the total weight to the search result, and then loops to block 802 to select the next search result. Inblock 814, the ranker component ranks the search results based on the total weights assigned to the search results, and completes. -
FIG. 9 is a graphical depiction of weight values as a function of proximity to an event, according to some embodiments. The graph illustrates the temporal weight value, which is the importance of an item of information (i.e., search result) to a user, relative to time. The relative importance of the item of information varies differently depending on the current time. In the graph depicted inFIG. 9 , T1 is the scheduled start time of the event related to the item of information, and T2 is the scheduled end time of the related event. In the time period before T1, the importance of the item of information to the user rises, for example, exponentially as depicted inFIG. 9 , but with a shallow upward curve as the current time approaches closer to T1. During the event, the importance of the item of information to the user remains relatively flat. During the time after the end of the event, the importance of the item of information to the user quickly drops, for example, exponentially as depicted inFIG. 9 , so that the importance diminishes to a value of zero in a very short amount of time. In some embodiments, as depicted inFIG. 10 , the importance of the item of information to the user during the event can rise to a peak at a point in time during the event, for example at time (T1+T2)/2, and decrease thereafter to the scheduled end of the event. -
FIG. 11 is a flow diagram that illustrates the processing of the ranker component to calculate a temporal weight of a search result, according to some embodiments. Inblock 1102, the ranker component identifies the event that corresponds to the search result. Inblock 1104, the ranker component determines the current time (i.e., the time at the present). Inblock 1106, the ranker component determines the time the identified event is scheduled to start. Inblock 1108, the ranker component determines the time the identified event is scheduled to end. Indecision block 1110, if the current time is before the time the event is scheduled to start, then the ranker component continues atblock 1112, else the ranker component continues atdecision block 1114. Inblock 1112, the ranker component calculates a temporal weight for the search result using Equation 3, and completes. Indecision block 1114, if the current time is between the time the event is scheduled to start and the time the event is scheduled to end, then the ranker component continues atblock 1116, else the ranker component continues atblock 1118. Inblock 1116, the ranker component calculates a temporal weight for the search result using Equation 4, and completes. Inblock 1118, the ranker component calculates a temporal weight for the search result using Equation 5, and completes. -
FIG. 12 is a flow diagram that illustrates the rendering information based on an event class corresponding to a user event, according to some embodiments. The information dissemination system creates event classes and disseminates information related to topics extracted from the event classes. Inblock 1202, a component of the information dissemination system creates event classes. Inblock 1204, the component identifies a user event to process. Inblock 1206, the component determines an event class that corresponds to the identified user event. Inblock 1208, the component processes the identified event class, and completes. -
FIG. 13 is a flow diagram that illustrates the creating of an event class, according to some embodiments. A component of the information dissemination system may create an event class in response to a user request to create an event class and based upon input provided by the user. Inblock 1302, the component creates an event class recognizer for the event class. Inblock 1304, the component creates a set of special topics for the event class. Inblock 1306, the component selects the next special topic. Indecision block 1308, if all of the special topics have already been selected, then the component completes, else the component continues atblock 1310. Inblock 1310, the component creates a set of slots for the special topic. Inblock 1312, the component creates a slot recognizer for each of the created slots. Inblock 1314, the component creates a set of actions for the special topic. Inblock 1316, the component creates a set of rules for digesting the action results to extract the relevant information from the action results, and then loops to block 1306 to select the next special topic. -
FIG. 14 is an example data listing that illustrates an event class, according to some embodiments. The listing illustrates a plurality of special topics contained in the event class, including a “Travel advisory topic”section 1402 for a travel advisory special topic. The travel advisory topic section contains a “Slots”section 1404, a “Recognizers”section 1406, and an “Action”section 1408. The slots section specifies the slots that have been defined for the special topic. For example, the slots “ToCity” and “ToCountry” have been defined for the travel advisory topic. The recognizers section specifies the restrictions on the values of the slots and how to determine the values for the slots. The action section specifies the actions that are to be taken for the special topic. For example, the actions may be specified in the form of parameterized URLs to be displayed or executed. The listing inFIG. 14 also illustrates other special topics which have not been illustrated to the same level of detail as the travel advisory special topic. Although not depicted, each of these special topics may also contain sections similar to those contained in the travel advisory special topic. -
FIG. 15 is a flow diagram that illustrates processing an event class, according to some embodiments. A component of the information dissemination system may process an event class upon identifying a user event to process. Inblock 1502, the component identifies the special topics that have been defined (i.e., created) for the event class. Inblock 1504, the component selects the next defined special topic. Indecision block 1506, if all of the defined special topics have already been selected, then the component completes, else the component continues atblock 1508. Inblock 1508, the component identifies the slots that have been defined for the special topic. Inblock 1510, the component selects the next defined slot. Indecision block 1512, if all of the defined slots have already been selected, then the component continues atblock 1516, else the component continues atblock 1514. Inblock 1514, the component uses the slot recognizer for the slot to identify the slot filler (and a value) for the slot, and then loops to block 1510 to select the next defined slot. In some embodiments, in instances where the component is unable to identify the slot filler, the component may request the user to provide the slot filler. Inblock 1516, the component identifies the actions that have been defined for the special topic. Inblock 1518, the component performs the actions. Inblock 1520, the component digests the action results, for example, to identify the relevant information. Inblock 1522, the component renders the digested action results to the user, and then loops to block 1504 to select the next defined special topic. - Although the subject matter has been described in language specific to structural features and/or methodological acts, it is to be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are disclosed as example forms of implementing the claims. Accordingly, the invention is not limited except as by the appended claims.
Claims (21)
1-20. (canceled)
21. A method, performed by a computer having a memory and a processor, for ranking search results, the method comprising:
for each of a plurality of search results,
identifying an event corresponding to the search result, wherein the identified event has an associated start time, and
determining a temporal weight for the search result based at least in part on a temporal proximity of a current time to the start time associated with the identified event; and
ordering the search results based at least in part on the determined temporal weights.
22. The method of claim 21 wherein the identified event has an associated end time and wherein determining the temporal weight is further based at least in part on a temporal proximity of the current time to the end time associated with the identified event.
23. The method of claim 21 , further comprising:
determining a static weight for each search result; and
ordering the search results based at least in part on the determined static weights.
24. The method of claim 23 wherein the determined static weight for each search result is based at least in part on an importance of the event corresponding to that search result.
25. The method of claim 23 wherein the determined static weight for each search result is based at least in part on an importance of the source of that search result.
26. The method of claim 21 wherein each of the identified events are associated with an activity scheduled by a calendaring application.
27. A computer-readable storage medium containing instructions for performing steps for generating a user profile, the steps comprising:
identifying information associated with a user;
extracting terms from the identified information associated with the user;
generating clusters of similar terms from the extracted terms;
identifying a plurality of events;
identifying a generated cluster related to the identified event;
identifying key terms for the identified cluster;
generating a query from the identified key terms; and
placing an indication of the generated query in a profile associated with the user.
28. The computer-readable storage medium of claim 27 , further comprising instructions for performing steps comprising:
obtaining search results by executing at least one generated query against at least one information source;
ranking the obtained search results; and
rendering the ranked search results to the user.
29. The computer-readable storage medium of claim 27 wherein the identified information includes email messages.
30. The computer-readable storage medium of claim 29 wherein the identified information includes World Wide Web usage information associated with the user.
31. The computer-readable storage medium of claim 30 wherein the identified information includes enterprise resource planning data associated with the user.
32. The computer-readable storage medium of claim 28 wherein an event of the plurality of events is identified from a calendar entry for a meeting with a start time and wherein the steps further comprise executing the query before the meeting and presenting results of the executed query to the user.
33. A system comprising:
a component configured to, for each of a plurality of search results,
identify an event corresponding to the search result, wherein the identified event has an associated start time, and
determine a temporal weight for the search result based at least in part on a temporal proximity of a current time to the start time associated with the identified event; and
a component configured to order the search results based at least in part on the determined temporal weights.
34. The system of claim 33 wherein the identified event has an associated end time and wherein the component configured to determine a temporal weights is further configured to determine the temporal weight based at least in part on a temporal proximity of the current time to the end time associated with the identified event.
35. The system of claim 34 , further comprising:
a component configured to determine an adaptive weight for each search result wherein the component configured to order the search results is further configured to order the search results based at least in part on the determined adaptive weights.
36. The system of claim 35 wherein the determined adaptive weight for each search result is based at least in part on preference information associated with a user.
37. The system of claim 36 wherein the determined static weight for each search result is based at least in part on feedback received from the user.
38. The system of claim 37 , further comprising:
a component that determines a static weight for each search result
wherein the component configured to order the search results is further configured to order the search results based at least in part on the determined static weights.
39. The system of claim 38 wherein each of the identified events are associated with an activity scheduled by a calendaring application.
40. The system of claim 39 wherein the search results are web pages.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US12/911,696 US20110040751A1 (en) | 2006-07-31 | 2010-10-25 | Temporal ranking of search results |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US11/461,383 US7849079B2 (en) | 2006-07-31 | 2006-07-31 | Temporal ranking of search results |
US12/911,696 US20110040751A1 (en) | 2006-07-31 | 2010-10-25 | Temporal ranking of search results |
Related Parent Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US11/461,383 Continuation US7849079B2 (en) | 2006-07-31 | 2006-07-31 | Temporal ranking of search results |
Publications (1)
Publication Number | Publication Date |
---|---|
US20110040751A1 true US20110040751A1 (en) | 2011-02-17 |
Family
ID=38987595
Family Applications (2)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US11/461,383 Expired - Fee Related US7849079B2 (en) | 2006-07-31 | 2006-07-31 | Temporal ranking of search results |
US12/911,696 Abandoned US20110040751A1 (en) | 2006-07-31 | 2010-10-25 | Temporal ranking of search results |
Family Applications Before (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US11/461,383 Expired - Fee Related US7849079B2 (en) | 2006-07-31 | 2006-07-31 | Temporal ranking of search results |
Country Status (11)
Country | Link |
---|---|
US (2) | US7849079B2 (en) |
EP (1) | EP2047389A4 (en) |
JP (1) | JP2009545810A (en) |
CN (1) | CN101496009A (en) |
AU (1) | AU2007281645B2 (en) |
BR (1) | BRPI0714327A2 (en) |
MX (1) | MX2009000591A (en) |
NO (1) | NO20085388L (en) |
RU (1) | RU2435213C2 (en) |
TW (1) | TW200817943A (en) |
WO (1) | WO2008016498A1 (en) |
Cited By (20)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20090240691A1 (en) * | 2008-03-24 | 2009-09-24 | Fujitsu Limited | Recording medium recording object contents search support program, object contents search support method, and object contents search support apparatus |
US20100131542A1 (en) * | 2008-11-26 | 2010-05-27 | James Paul Schneider | Suggested websites |
US20100223578A1 (en) * | 2009-03-02 | 2010-09-02 | Bernardo Huberman | Populating variable content slots on web pages |
US8166026B1 (en) * | 2006-12-26 | 2012-04-24 | uAffect.org LLC | User-centric, user-weighted method and apparatus for improving relevance and analysis of information sharing and searching |
US8682895B1 (en) * | 2011-03-31 | 2014-03-25 | Twitter, Inc. | Content resonance |
US9105029B2 (en) * | 2011-09-19 | 2015-08-11 | Ebay Inc. | Search system utilizing purchase history |
US9117227B1 (en) * | 2011-03-31 | 2015-08-25 | Twitter, Inc. | Temporal features in a messaging platform |
US9319359B1 (en) * | 2011-03-31 | 2016-04-19 | Twitter, Inc. | Promoting content in a real-time messaging platform |
US20180211184A1 (en) * | 2017-01-20 | 2018-07-26 | Mehdi El Mountassir | Method and system for virtual connection between traveler and local concierge based on similar interests background |
US10248667B1 (en) | 2013-03-15 | 2019-04-02 | Twitter, Inc. | Pre-filtering in a messaging platform |
US10348792B2 (en) | 2005-12-14 | 2019-07-09 | Facebook, Inc. | Dynamically updating media content for display to a user of a social network environment based on user interactions |
US10650408B1 (en) | 2013-03-15 | 2020-05-12 | Twitter, Inc. | Budget smoothing in a messaging platform |
US10769677B1 (en) * | 2011-03-31 | 2020-09-08 | Twitter, Inc. | Temporal features in a messaging platform |
US20210224346A1 (en) | 2018-04-20 | 2021-07-22 | Facebook, Inc. | Engaging Users by Personalized Composing-Content Recommendation |
US11263255B2 (en) * | 2017-01-20 | 2022-03-01 | Twitter, Inc. | Content carousel in a social media timeline |
US11307880B2 (en) | 2018-04-20 | 2022-04-19 | Meta Platforms, Inc. | Assisting users with personalized and contextual communication content |
US11443348B1 (en) | 2016-01-05 | 2022-09-13 | Twitter, Inc. | Conversational targeted content |
US11676220B2 (en) | 2018-04-20 | 2023-06-13 | Meta Platforms, Inc. | Processing multimodal user input for assistant systems |
US11715042B1 (en) | 2018-04-20 | 2023-08-01 | Meta Platforms Technologies, Llc | Interpretability of deep reinforcement learning models in assistant systems |
US11886473B2 (en) | 2018-04-20 | 2024-01-30 | Meta Platforms, Inc. | Intent identification for agent matching by assistant systems |
Families Citing this family (64)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8402094B2 (en) * | 2006-08-11 | 2013-03-19 | Facebook, Inc. | Providing a newsfeed based on user affinity for entities and monitored actions in a social network environment |
US7685199B2 (en) * | 2006-07-31 | 2010-03-23 | Microsoft Corporation | Presenting information related to topics extracted from event classes |
US7577718B2 (en) * | 2006-07-31 | 2009-08-18 | Microsoft Corporation | Adaptive dissemination of personalized and contextually relevant information |
US7783630B1 (en) * | 2007-06-29 | 2010-08-24 | Emc Corporation | Tuning of relevancy ranking for federated search |
US7783620B1 (en) * | 2007-06-29 | 2010-08-24 | Emc Corporation | Relevancy scoring using query structure and data structure for federated search |
US8442969B2 (en) * | 2007-08-14 | 2013-05-14 | John Nicholas Gross | Location based news and search engine |
US7797314B2 (en) * | 2007-12-31 | 2010-09-14 | International Business Machines Corporation | Adaptive searching |
US8626575B1 (en) | 2008-03-20 | 2014-01-07 | Google Inc. | Targeted advertising using temporal analysis of user specific data |
US8326827B2 (en) * | 2008-04-14 | 2012-12-04 | Magic Network Sarl | Network peer-to-peer goods and services delivery system and method for ranking peers by degrees of association |
US20100082639A1 (en) * | 2008-09-30 | 2010-04-01 | Microsoft Corporation | Processing maximum likelihood for listwise rankings |
US8468153B2 (en) * | 2009-01-21 | 2013-06-18 | Recorded Future, Inc. | Information service for facts extracted from differing sources on a wide area network |
US20110099488A1 (en) * | 2009-10-26 | 2011-04-28 | Verizon Patent And Licensing Inc. | Method and apparatus for presenting video assets |
TWI616761B (en) * | 2010-03-09 | 2018-03-01 | Alibaba Group Holding Ltd | Information matching method and system applied to e-commerce website |
US9785987B2 (en) | 2010-04-22 | 2017-10-10 | Microsoft Technology Licensing, Llc | User interface for information presentation system |
CN102236678A (en) * | 2010-04-26 | 2011-11-09 | 腾讯科技(深圳)有限公司 | Method and system for managing dynamic information of friends, and server for managing information of friends |
US20110264642A1 (en) * | 2010-04-26 | 2011-10-27 | Microsoft Corporation | Dynamic computation engine in search stack |
US9215420B2 (en) | 2010-06-01 | 2015-12-15 | Comcast Cable Communications, Llc | Ranking search results |
US8786597B2 (en) | 2010-06-30 | 2014-07-22 | International Business Machines Corporation | Management of a history of a meeting |
US9043296B2 (en) | 2010-07-30 | 2015-05-26 | Microsoft Technology Licensing, Llc | System of providing suggestions based on accessible and contextual information |
US8687941B2 (en) | 2010-10-29 | 2014-04-01 | International Business Machines Corporation | Automatic static video summarization |
US9208230B2 (en) * | 2010-10-29 | 2015-12-08 | Google Inc. | Enriching search results |
US8527483B2 (en) * | 2011-02-04 | 2013-09-03 | Mikko VÄÄNÄNEN | Method and means for browsing by walking |
TW201235867A (en) * | 2011-02-18 | 2012-09-01 | Hon Hai Prec Ind Co Ltd | System and method for searching related terms |
US8630860B1 (en) * | 2011-03-03 | 2014-01-14 | Nuance Communications, Inc. | Speaker and call characteristic sensitive open voice search |
CN102117332A (en) * | 2011-03-10 | 2011-07-06 | 辜进荣 | Given time-based searching method |
JP2012226649A (en) * | 2011-04-21 | 2012-11-15 | Sony Corp | Information processing apparatus, information processing method, and program |
US20130106894A1 (en) | 2011-10-31 | 2013-05-02 | Elwha LLC, a limited liability company of the State of Delaware | Context-sensitive query enrichment |
US20130132138A1 (en) * | 2011-11-23 | 2013-05-23 | International Business Machines Corporation | Identifying influence paths and expertise network in an enterprise using meeting provenance data |
TWI477996B (en) * | 2011-11-29 | 2015-03-21 | Iq Technology Inc | Method of analyzing personalized input automatically |
CN102419773B (en) * | 2011-12-13 | 2014-09-03 | 百度在线网络技术(北京)有限公司 | Method, device and equipment used for sequencing resource items |
CN102426610B (en) * | 2012-01-13 | 2014-05-07 | 中国科学院计算技术研究所 | Microblog rank searching method and microblog searching engine |
US8818176B2 (en) * | 2012-02-21 | 2014-08-26 | Avaya Inc. | System and method for aligning tags to specific video frames |
US9311650B2 (en) | 2012-02-22 | 2016-04-12 | Alibaba Group Holding Limited | Determining search result rankings based on trust level values associated with sellers |
CN107066509B (en) * | 2012-04-26 | 2020-12-04 | 阿里巴巴集团控股有限公司 | Information providing method, processing server and merge server |
US8914452B2 (en) | 2012-05-31 | 2014-12-16 | International Business Machines Corporation | Automatically generating a personalized digest of meetings |
US10664768B2 (en) * | 2012-06-13 | 2020-05-26 | Expedia, Inc. | Travel advisory notifications |
JP5958134B2 (en) * | 2012-07-18 | 2016-07-27 | 株式会社リコー | SEARCH DEVICE, SEARCH PROGRAM, AND SEARCH SYSTEM |
US9237386B2 (en) | 2012-08-31 | 2016-01-12 | Google Inc. | Aiding discovery of program content by providing deeplinks into most interesting moments via social media |
EP2897057A4 (en) * | 2012-09-13 | 2016-05-04 | Ntt Docomo Inc | User interface device, search method, and program |
US10127602B2 (en) | 2012-11-06 | 2018-11-13 | Ebay Inc. | Systems and methods for transient local commerce search |
US9401947B1 (en) | 2013-02-08 | 2016-07-26 | Google Inc. | Methods, systems, and media for presenting comments based on correlation with content |
US9213744B2 (en) * | 2013-03-15 | 2015-12-15 | Ifwe Inc. | Perturbing search results |
RU2543315C2 (en) | 2013-03-22 | 2015-02-27 | Федеральное государственное автономное образовательное учреждение высшего профессионального образования "Национальный исследовательский университет "Высшая школа экономики" | Method of selecting effective versions in search and recommendation systems (versions) |
CN103226858B (en) * | 2013-04-02 | 2016-08-24 | 深圳市文鼎创数据科技有限公司 | The processing method and processing device of Bluetooth pairing information |
US9734208B1 (en) * | 2013-05-13 | 2017-08-15 | Audible, Inc. | Knowledge sharing based on meeting information |
US11263221B2 (en) | 2013-05-29 | 2022-03-01 | Microsoft Technology Licensing, Llc | Search result contexts for application launch |
US10430418B2 (en) | 2013-05-29 | 2019-10-01 | Microsoft Technology Licensing, Llc | Context-based actions from a source application |
CN103646106B (en) * | 2013-12-23 | 2016-05-25 | 山东大学 | A kind of Web theme sort method of content-based similitude |
US9454621B2 (en) * | 2013-12-31 | 2016-09-27 | Google Inc. | Surfacing navigational search results |
CN103942264B (en) * | 2014-03-26 | 2017-09-01 | 北京奇虎科技有限公司 | The method and apparatus for pushing the webpage comprising news information |
US11463541B2 (en) * | 2014-08-02 | 2022-10-04 | Google Llc | Providing content based on event related information |
US10229164B1 (en) | 2014-08-02 | 2019-03-12 | Google Llc | Adjusting a relevancy score of a keyword cluster—time period—event category combination based on event related information |
US9843649B1 (en) | 2014-08-02 | 2017-12-12 | Google Llc | Providing content based on event related information |
US9779144B1 (en) | 2014-08-02 | 2017-10-03 | Google Inc. | Identifying a level of relevancy of a keyword cluster related to an event category for a given time period relative to the event |
RU2580516C2 (en) * | 2014-08-19 | 2016-04-10 | Общество С Ограниченной Ответственностью "Яндекс" | Method of generating customised ranking model, method of generating ranking model, electronic device and server |
US11093971B2 (en) * | 2015-02-27 | 2021-08-17 | Keypoint Technologies India Pvt Ltd. | Contextual discovery |
JP6593873B2 (en) * | 2015-10-07 | 2019-10-23 | Necソリューションイノベータ株式会社 | Information analysis apparatus and information analysis method |
US10389543B2 (en) | 2015-12-31 | 2019-08-20 | Microsoft Technology Licensing, Llc | Starting meeting using natural user input |
CN107809363B (en) * | 2016-08-30 | 2021-07-27 | 腾讯科技(深圳)有限公司 | Network information propagation control method and device |
US10691734B2 (en) * | 2017-11-21 | 2020-06-23 | International Business Machines Corporation | Searching multilingual documents based on document structure extraction |
CN110209918A (en) * | 2018-02-13 | 2019-09-06 | 阿里巴巴集团控股有限公司 | A kind of text handling method based on date event, device and electronic equipment |
WO2020102033A2 (en) * | 2018-11-12 | 2020-05-22 | Nant Holdings Ip, Llc | Curation and provision of digital content |
CN112491925A (en) * | 2020-12-10 | 2021-03-12 | 北京冠程科技有限公司 | Method, system and electronic equipment for acquiring network security event according to time node |
CN112927770B (en) * | 2021-04-12 | 2023-09-08 | 徐州市通用科技有限公司 | Medical data sharing method and system |
Citations (20)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5732399A (en) * | 1995-06-07 | 1998-03-24 | Sun Microsystems, Inc. | Method and apparatus for associating calendar events with contextual information and other calendar events |
US5841965A (en) * | 1994-05-16 | 1998-11-24 | Ricoh Company, Ltd. | System and method for automatically determining test point for DC parametric test |
US20020152185A1 (en) * | 2001-01-03 | 2002-10-17 | Sasken Communication Technologies Limited | Method of network modeling and predictive event-correlation in a communication system by the use of contextual fuzzy cognitive maps |
US20030037063A1 (en) * | 2001-08-10 | 2003-02-20 | Qlinx | Method and system for dynamic risk assessment, risk monitoring, and caseload management |
US20030135490A1 (en) * | 2002-01-15 | 2003-07-17 | Barrett Michael E. | Enhanced popularity ranking |
US20030158855A1 (en) * | 2002-02-20 | 2003-08-21 | Farnham Shelly D. | Computer system architecture for automatic context associations |
US20030182263A1 (en) * | 2000-01-28 | 2003-09-25 | Augustine Andrew W. | Investor relations event notification system and method |
US6640230B1 (en) * | 2000-09-27 | 2003-10-28 | International Business Machines Corporation | Calendar-driven application technique for preparing responses to incoming events |
US20040044658A1 (en) * | 2000-11-20 | 2004-03-04 | Crabtree Ian B | Information provider |
US20040254998A1 (en) * | 2000-06-17 | 2004-12-16 | Microsoft Corporation | When-free messaging |
US20040267730A1 (en) * | 2003-06-26 | 2004-12-30 | Microsoft Corporation | Systems and methods for performing background queries from content and activity |
US20050114317A1 (en) * | 2003-11-26 | 2005-05-26 | Bhide Manish A. | Ordering of web search results |
US20050234848A1 (en) * | 2004-03-31 | 2005-10-20 | Lawrence Stephen R | Methods and systems for information capture and retrieval |
US20060010217A1 (en) * | 2004-06-04 | 2006-01-12 | Business Instruments Corp. | System and method for dynamic adaptive user-based prioritization and display of electronic messages |
US20060010023A1 (en) * | 2000-10-02 | 2006-01-12 | On Vantage, Inc. | System, method and computer program product for managing meeting planning operations |
WO2006014824A1 (en) * | 2004-07-26 | 2006-02-09 | Wireless 5Th Dimensional Networking, Inc. | Context-based search engine residing on a network |
US20060047635A1 (en) * | 2004-08-26 | 2006-03-02 | International Business Machines Corporation | Method of generating a context-inferenced search query and of sorting a result of the query |
US20060136451A1 (en) * | 2004-12-22 | 2006-06-22 | Mikhail Denissov | Methods and systems for applying attention strength, activation scores and co-occurrence statistics in information management |
US20060287993A1 (en) * | 2005-06-21 | 2006-12-21 | Microsoft Corporation | High scale adaptive search systems and methods |
US20070027852A1 (en) * | 2005-07-29 | 2007-02-01 | Microsoft Corporation | Smart search for accessing options |
Family Cites Families (77)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US4965763A (en) * | 1987-03-03 | 1990-10-23 | International Business Machines Corporation | Computer method for automatic extraction of commonly specified information from business correspondence |
US6321158B1 (en) * | 1994-06-24 | 2001-11-20 | Delorme Publishing Company | Integrated routing/mapping information |
US6460036B1 (en) * | 1994-11-29 | 2002-10-01 | Pinpoint Incorporated | System and method for providing customized electronic newspapers and target advertisements |
DE69531599T2 (en) * | 1994-12-20 | 2004-06-24 | Sun Microsystems, Inc., Mountain View | Method and device for finding and obtaining personalized information |
US5685000A (en) * | 1995-01-04 | 1997-11-04 | U S West Technologies, Inc. | Method for providing a linguistically competent dialogue with a computerized service representative |
US6442527B1 (en) * | 1995-03-17 | 2002-08-27 | Kathy A. Worthington | System and method for personalized and customized time management |
US6163274A (en) * | 1997-09-04 | 2000-12-19 | Ncr Corporation | Remotely updatable PDA |
US6594682B2 (en) * | 1997-10-28 | 2003-07-15 | Microsoft Corporation | Client-side system for scheduling delivery of web content and locally managing the web content |
US6266668B1 (en) * | 1998-08-04 | 2001-07-24 | Dryken Technologies, Inc. | System and method for dynamic data-mining and on-line communication of customized information |
US6845370B2 (en) * | 1998-11-12 | 2005-01-18 | Accenture Llp | Advanced information gathering for targeted activities |
US6802042B2 (en) * | 1999-06-01 | 2004-10-05 | Yodlee.Com, Inc. | Method and apparatus for providing calculated and solution-oriented personalized summary-reports to a user through a single user-interface |
US6920616B1 (en) * | 1998-12-18 | 2005-07-19 | Tangis Corporation | Interface for exchanging context data |
US6848104B1 (en) * | 1998-12-21 | 2005-01-25 | Koninklijke Philips Electronics N.V. | Clustering of task-associated objects for effecting tasks among a system and its environmental devices |
US6317718B1 (en) | 1999-02-26 | 2001-11-13 | Accenture Properties (2) B.V. | System, method and article of manufacture for location-based filtering for shopping agent in the physical world |
US6202062B1 (en) * | 1999-02-26 | 2001-03-13 | Ac Properties B.V. | System, method and article of manufacture for creating a filtered information summary based on multiple profiles of each single user |
US6327590B1 (en) * | 1999-05-05 | 2001-12-04 | Xerox Corporation | System and method for collaborative ranking of search results employing user and group profiles derived from document collection content analysis |
IL147386A0 (en) * | 1999-06-30 | 2002-08-14 | Silverbrook Res Pty Ltd | Method and system for banking |
US6873982B1 (en) * | 1999-07-16 | 2005-03-29 | International Business Machines Corporation | Ordering of database search results based on user feedback |
US6675350B1 (en) * | 1999-11-04 | 2004-01-06 | International Business Machines Corporation | System for collecting and displaying summary information from disparate sources |
US6963867B2 (en) * | 1999-12-08 | 2005-11-08 | A9.Com, Inc. | Search query processing to provide category-ranked presentation of search results |
US6834195B2 (en) * | 2000-04-04 | 2004-12-21 | Carl Brock Brandenberg | Method and apparatus for scheduling presentation of digital content on a personal communication device |
US6778973B2 (en) * | 2000-01-12 | 2004-08-17 | Duane Harlan | Method for selection of events based on proximity |
US6546388B1 (en) * | 2000-01-14 | 2003-04-08 | International Business Machines Corporation | Metadata search results ranking system |
US6564210B1 (en) * | 2000-03-27 | 2003-05-13 | Virtual Self Ltd. | System and method for searching databases employing user profiles |
US6553378B1 (en) * | 2000-03-31 | 2003-04-22 | Network Associates, Inc. | System and process for reporting network events with a plurality of hierarchically-structured databases in a distributed computing environment |
US7599847B2 (en) * | 2000-06-09 | 2009-10-06 | Airport America | Automated internet based interactive travel planning and management system |
US7003513B2 (en) * | 2000-07-04 | 2006-02-21 | International Business Machines Corporation | Method and system of weighted context feedback for result improvement in information retrieval |
DE60010087T2 (en) * | 2000-10-16 | 2005-03-24 | Alcatel | Method and device for transmitting an information message with a customized content to a user or a group of users of a communication terminal |
US7721210B2 (en) * | 2000-11-29 | 2010-05-18 | Nokia Corporation | Electronic calendar system |
US6907423B2 (en) * | 2001-01-04 | 2005-06-14 | Sun Microsystems, Inc. | Search engine interface and method of controlling client searches |
US7003730B2 (en) * | 2002-03-08 | 2006-02-21 | International Business Machines Corporation | Graphical user interface to build event-based dynamic searches or queries using event profiles |
KR100988153B1 (en) | 2002-03-29 | 2010-10-18 | 소니 주식회사 | Information search system, information processing apparatus and method, and information search apparatus and method |
US20040054550A1 (en) * | 2002-04-04 | 2004-03-18 | James Cole | System and method for the distribution of information during irregular operations |
US7146359B2 (en) * | 2002-05-03 | 2006-12-05 | Hewlett-Packard Development Company, L.P. | Method and system for filtering content in a discovered topic |
US6920459B2 (en) * | 2002-05-07 | 2005-07-19 | Zycus Infotech Pvt Ltd. | System and method for context based searching of electronic catalog database, aided with graphical feedback to the user |
US7177859B2 (en) * | 2002-06-26 | 2007-02-13 | Microsoft Corporation | Programming model for subscription services |
US20040002988A1 (en) * | 2002-06-26 | 2004-01-01 | Praveen Seshadri | System and method for modeling subscriptions and subscribers as data |
JP4005425B2 (en) * | 2002-06-28 | 2007-11-07 | 富士通株式会社 | Search result ranking change processing program, search result ranking change processing program recording medium, and content search processing method |
US20040128295A1 (en) * | 2002-12-27 | 2004-07-01 | International Business Machines Corporation | Data structure depicting an event for use in a computer implemented situation manager and method and system for use therewith |
EP1586058A1 (en) * | 2003-01-24 | 2005-10-19 | BRITISH TELECOMMUNICATIONS public limited company | Searching apparatus and methods |
US20040181604A1 (en) * | 2003-03-13 | 2004-09-16 | Immonen Pekka S. | System and method for enhancing the relevance of push-based content |
US20050015307A1 (en) * | 2003-04-28 | 2005-01-20 | Simpson Todd Garrett | Method and system of providing location sensitive business information to customers |
US7162473B2 (en) * | 2003-06-26 | 2007-01-09 | Microsoft Corporation | Method and system for usage analyzer that determines user accessed sources, indexes data subsets, and associated metadata, processing implicit queries based on potential interest to users |
US7836010B2 (en) * | 2003-07-30 | 2010-11-16 | Northwestern University | Method and system for assessing relevant properties of work contexts for use by information services |
JP4349875B2 (en) * | 2003-09-19 | 2009-10-21 | 株式会社リコー | Document filtering apparatus, document filtering method, and document filtering program |
US20050071328A1 (en) * | 2003-09-30 | 2005-03-31 | Lawrence Stephen R. | Personalization of web search |
US7693827B2 (en) * | 2003-09-30 | 2010-04-06 | Google Inc. | Personalization of placed content ordering in search results |
US8489769B2 (en) * | 2003-10-02 | 2013-07-16 | Accenture Global Services Limited | Intelligent collaborative expression in support of socialization of devices |
US8024335B2 (en) * | 2004-05-03 | 2011-09-20 | Microsoft Corporation | System and method for dynamically generating a selectable search extension |
WO2005045690A1 (en) | 2003-11-04 | 2005-05-19 | Porto Ranelli, Sa | Method and system for contextual activation of web services |
CN100495392C (en) * | 2003-12-29 | 2009-06-03 | 西安迪戈科技有限责任公司 | Intelligent search method |
KR20040091532A (en) | 2004-01-16 | 2004-10-28 | 엔에이치엔(주) | A method of introducing advertisements and providing the advertisements by using access intentions of internet users and a system thereof |
US7392278B2 (en) * | 2004-01-23 | 2008-06-24 | Microsoft Corporation | Building and using subwebs for focused search |
US20050165762A1 (en) * | 2004-01-26 | 2005-07-28 | Thinkbig, Inc., A California Corporation | User event matching system and method |
US10417298B2 (en) * | 2004-12-02 | 2019-09-17 | Insignio Technologies, Inc. | Personalized content processing and delivery system and media |
US7716223B2 (en) * | 2004-03-29 | 2010-05-11 | Google Inc. | Variable personalization of search results in a search engine |
US7664734B2 (en) * | 2004-03-31 | 2010-02-16 | Google Inc. | Systems and methods for generating multiple implicit search queries |
US9009153B2 (en) * | 2004-03-31 | 2015-04-14 | Google Inc. | Systems and methods for identifying a named entity |
US7941439B1 (en) * | 2004-03-31 | 2011-05-10 | Google Inc. | Methods and systems for information capture |
US8078607B2 (en) * | 2006-03-30 | 2011-12-13 | Google Inc. | Generating website profiles based on queries from webistes and user activities on the search results |
US7617176B2 (en) * | 2004-07-13 | 2009-11-10 | Microsoft Corporation | Query-based snippet clustering for search result grouping |
US7487072B2 (en) * | 2004-08-04 | 2009-02-03 | International Business Machines Corporation | Method and system for querying multimedia data where adjusting the conversion of the current portion of the multimedia data signal based on the comparing at least one set of confidence values to the threshold |
US20060047551A1 (en) * | 2004-08-26 | 2006-03-02 | Sandra Cotten | System and method for staffing promotional events with qualified event personnel |
US8239375B2 (en) * | 2004-08-31 | 2012-08-07 | Research In Motion Limited | Method of searching for personal information management (PIM) information and handheld electronic device employing the same |
WO2006034038A2 (en) * | 2004-09-17 | 2006-03-30 | Become, Inc. | Systems and methods of retrieving topic specific information |
US7707167B2 (en) | 2004-09-20 | 2010-04-27 | Microsoft Corporation | Method, system, and apparatus for creating a knowledge interchange profile |
US7583646B2 (en) * | 2004-10-14 | 2009-09-01 | Alcatel-Lucent Usa Inc. | Method and apparatus for facilitating interaction between a home subscriber server (HSS) and a home location register (HLR) in a legacy network |
US7392169B2 (en) * | 2004-10-21 | 2008-06-24 | International Business Machines Corporation | Method, system and program product for defining and recording minimum and maximum event counts of a simulation utilizing a high level language |
US20060090184A1 (en) * | 2004-10-26 | 2006-04-27 | David Zito | System and method for presenting information |
US20060106768A1 (en) * | 2004-11-12 | 2006-05-18 | A W Seabiscuit B.V. | Method and website for making travel plans |
US7523137B2 (en) * | 2005-04-08 | 2009-04-21 | Accenture Global Services Gmbh | Model-driven event detection, implication, and reporting system |
US8229914B2 (en) * | 2005-09-14 | 2012-07-24 | Jumptap, Inc. | Mobile content spidering and compatibility determination |
US7734632B2 (en) * | 2005-10-28 | 2010-06-08 | Disney Enterprises, Inc. | System and method for targeted ad delivery |
US7734593B2 (en) * | 2005-11-28 | 2010-06-08 | Commvault Systems, Inc. | Systems and methods for classifying and transferring information in a storage network |
US7904448B2 (en) * | 2006-03-29 | 2011-03-08 | Yahoo! Inc. | Incremental update of long-term and short-term user profile scores in a behavioral targeting system |
US7577718B2 (en) * | 2006-07-31 | 2009-08-18 | Microsoft Corporation | Adaptive dissemination of personalized and contextually relevant information |
US7685199B2 (en) * | 2006-07-31 | 2010-03-23 | Microsoft Corporation | Presenting information related to topics extracted from event classes |
-
2006
- 2006-07-31 US US11/461,383 patent/US7849079B2/en not_active Expired - Fee Related
-
2007
- 2007-06-28 TW TW096123588A patent/TW200817943A/en unknown
- 2007-07-20 CN CNA2007800279427A patent/CN101496009A/en active Pending
- 2007-07-20 MX MX2009000591A patent/MX2009000591A/en active IP Right Grant
- 2007-07-20 BR BRPI0714327-3A patent/BRPI0714327A2/en not_active IP Right Cessation
- 2007-07-20 WO PCT/US2007/016507 patent/WO2008016498A1/en active Application Filing
- 2007-07-20 EP EP07836177A patent/EP2047389A4/en not_active Withdrawn
- 2007-07-20 AU AU2007281645A patent/AU2007281645B2/en not_active Ceased
- 2007-07-20 RU RU2009103144/08A patent/RU2435213C2/en not_active IP Right Cessation
- 2007-07-20 JP JP2009522780A patent/JP2009545810A/en not_active Withdrawn
-
2008
- 2008-12-29 NO NO20085388A patent/NO20085388L/en not_active Application Discontinuation
-
2010
- 2010-10-25 US US12/911,696 patent/US20110040751A1/en not_active Abandoned
Patent Citations (20)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5841965A (en) * | 1994-05-16 | 1998-11-24 | Ricoh Company, Ltd. | System and method for automatically determining test point for DC parametric test |
US5732399A (en) * | 1995-06-07 | 1998-03-24 | Sun Microsystems, Inc. | Method and apparatus for associating calendar events with contextual information and other calendar events |
US20030182263A1 (en) * | 2000-01-28 | 2003-09-25 | Augustine Andrew W. | Investor relations event notification system and method |
US20040254998A1 (en) * | 2000-06-17 | 2004-12-16 | Microsoft Corporation | When-free messaging |
US6640230B1 (en) * | 2000-09-27 | 2003-10-28 | International Business Machines Corporation | Calendar-driven application technique for preparing responses to incoming events |
US20060010023A1 (en) * | 2000-10-02 | 2006-01-12 | On Vantage, Inc. | System, method and computer program product for managing meeting planning operations |
US20040044658A1 (en) * | 2000-11-20 | 2004-03-04 | Crabtree Ian B | Information provider |
US20020152185A1 (en) * | 2001-01-03 | 2002-10-17 | Sasken Communication Technologies Limited | Method of network modeling and predictive event-correlation in a communication system by the use of contextual fuzzy cognitive maps |
US20030037063A1 (en) * | 2001-08-10 | 2003-02-20 | Qlinx | Method and system for dynamic risk assessment, risk monitoring, and caseload management |
US20030135490A1 (en) * | 2002-01-15 | 2003-07-17 | Barrett Michael E. | Enhanced popularity ranking |
US20030158855A1 (en) * | 2002-02-20 | 2003-08-21 | Farnham Shelly D. | Computer system architecture for automatic context associations |
US20040267730A1 (en) * | 2003-06-26 | 2004-12-30 | Microsoft Corporation | Systems and methods for performing background queries from content and activity |
US20050114317A1 (en) * | 2003-11-26 | 2005-05-26 | Bhide Manish A. | Ordering of web search results |
US20050234848A1 (en) * | 2004-03-31 | 2005-10-20 | Lawrence Stephen R | Methods and systems for information capture and retrieval |
US20060010217A1 (en) * | 2004-06-04 | 2006-01-12 | Business Instruments Corp. | System and method for dynamic adaptive user-based prioritization and display of electronic messages |
WO2006014824A1 (en) * | 2004-07-26 | 2006-02-09 | Wireless 5Th Dimensional Networking, Inc. | Context-based search engine residing on a network |
US20060047635A1 (en) * | 2004-08-26 | 2006-03-02 | International Business Machines Corporation | Method of generating a context-inferenced search query and of sorting a result of the query |
US20060136451A1 (en) * | 2004-12-22 | 2006-06-22 | Mikhail Denissov | Methods and systems for applying attention strength, activation scores and co-occurrence statistics in information management |
US20060287993A1 (en) * | 2005-06-21 | 2006-12-21 | Microsoft Corporation | High scale adaptive search systems and methods |
US20070027852A1 (en) * | 2005-07-29 | 2007-02-01 | Microsoft Corporation | Smart search for accessing options |
Cited By (62)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US10348792B2 (en) | 2005-12-14 | 2019-07-09 | Facebook, Inc. | Dynamically updating media content for display to a user of a social network environment based on user interactions |
US10078696B1 (en) * | 2006-12-26 | 2018-09-18 | Paul Jerome Sadler | Relevant social searching and user centric data analysis via user and peer group parameters via a dynamic interface |
US8166026B1 (en) * | 2006-12-26 | 2012-04-24 | uAffect.org LLC | User-centric, user-weighted method and apparatus for improving relevance and analysis of information sharing and searching |
US8244704B2 (en) * | 2008-03-24 | 2012-08-14 | Fujitsu Limited | Recording medium recording object contents search support program, object contents search support method, and object contents search support apparatus |
US20090240691A1 (en) * | 2008-03-24 | 2009-09-24 | Fujitsu Limited | Recording medium recording object contents search support program, object contents search support method, and object contents search support apparatus |
US20100131542A1 (en) * | 2008-11-26 | 2010-05-27 | James Paul Schneider | Suggested websites |
US9053197B2 (en) * | 2008-11-26 | 2015-06-09 | Red Hat, Inc. | Suggesting websites |
US20100223578A1 (en) * | 2009-03-02 | 2010-09-02 | Bernardo Huberman | Populating variable content slots on web pages |
US8566332B2 (en) * | 2009-03-02 | 2013-10-22 | Hewlett-Packard Development Company, L.P. | Populating variable content slots on web pages |
US8682895B1 (en) * | 2011-03-31 | 2014-03-25 | Twitter, Inc. | Content resonance |
US9892431B1 (en) * | 2011-03-31 | 2018-02-13 | Twitter, Inc. | Temporal features in a messaging platform |
US9298812B1 (en) * | 2011-03-31 | 2016-03-29 | Twitter, Inc. | Content resonance |
US9319359B1 (en) * | 2011-03-31 | 2016-04-19 | Twitter, Inc. | Promoting content in a real-time messaging platform |
US9454771B1 (en) * | 2011-03-31 | 2016-09-27 | Twitter, Inc. | Temporal features in a messaging platform |
US20160314498A1 (en) * | 2011-03-31 | 2016-10-27 | Twitter, Inc. | Promoting content in a real-time messaging platform |
US9524321B2 (en) | 2011-03-31 | 2016-12-20 | Twitter, Inc. | Content resonance |
US10803492B2 (en) | 2011-03-31 | 2020-10-13 | Twitter, Inc. | Promoting content in a real-time messaging platform |
US10970312B2 (en) | 2011-03-31 | 2021-04-06 | Twitter, Inc. | Content resonance |
US9117227B1 (en) * | 2011-03-31 | 2015-08-25 | Twitter, Inc. | Temporal features in a messaging platform |
US10146855B2 (en) | 2011-03-31 | 2018-12-04 | Twitter, Inc. | Content resonance |
US10163133B2 (en) * | 2011-03-31 | 2018-12-25 | Twitter, Inc. | Promoting content in a real-time messaging platform |
US10769677B1 (en) * | 2011-03-31 | 2020-09-08 | Twitter, Inc. | Temporal features in a messaging platform |
US9105029B2 (en) * | 2011-09-19 | 2015-08-11 | Ebay Inc. | Search system utilizing purchase history |
US20150310120A1 (en) * | 2011-09-19 | 2015-10-29 | Paypal, Inc. | Search system utilzing purchase history |
US10963922B1 (en) | 2013-03-15 | 2021-03-30 | Twitter, Inc. | Campaign goal setting in a messaging platform |
US10650408B1 (en) | 2013-03-15 | 2020-05-12 | Twitter, Inc. | Budget smoothing in a messaging platform |
US10769661B1 (en) | 2013-03-15 | 2020-09-08 | Twitter, Inc. | Real time messaging platform |
US10600080B1 (en) | 2013-03-15 | 2020-03-24 | Twitter, Inc. | Overspend control in a messaging platform |
US10248667B1 (en) | 2013-03-15 | 2019-04-02 | Twitter, Inc. | Pre-filtering in a messaging platform |
US11157464B1 (en) | 2013-03-15 | 2021-10-26 | Twitter, Inc. | Pre-filtering of candidate messages for message streams in a messaging platform |
US11216841B1 (en) | 2013-03-15 | 2022-01-04 | Twitter, Inc. | Real time messaging platform |
US11409717B1 (en) | 2013-03-15 | 2022-08-09 | Twitter, Inc. | Overspend control in a messaging platform |
US10692114B1 (en) | 2013-03-15 | 2020-06-23 | Twitter, Inc. | Exploration in a real time messaging platform |
US11288702B1 (en) | 2013-03-15 | 2022-03-29 | Twitter, Inc. | Exploration in a real time messaging platform |
US11443348B1 (en) | 2016-01-05 | 2022-09-13 | Twitter, Inc. | Conversational targeted content |
US11263255B2 (en) * | 2017-01-20 | 2022-03-01 | Twitter, Inc. | Content carousel in a social media timeline |
US20180211184A1 (en) * | 2017-01-20 | 2018-07-26 | Mehdi El Mountassir | Method and system for virtual connection between traveler and local concierge based on similar interests background |
US20220166733A1 (en) * | 2018-04-20 | 2022-05-26 | Meta Platforms, Inc. | Predictive Injection of Conversation Fillers for Assistant Systems |
US20230186618A1 (en) | 2018-04-20 | 2023-06-15 | Meta Platforms, Inc. | Generating Multi-Perspective Responses by Assistant Systems |
US11249773B2 (en) | 2018-04-20 | 2022-02-15 | Facebook Technologies, Llc. | Auto-completion for gesture-input in assistant systems |
US11301521B1 (en) | 2018-04-20 | 2022-04-12 | Meta Platforms, Inc. | Suggestions for fallback social contacts for assistant systems |
US11307880B2 (en) | 2018-04-20 | 2022-04-19 | Meta Platforms, Inc. | Assisting users with personalized and contextual communication content |
US11308169B1 (en) | 2018-04-20 | 2022-04-19 | Meta Platforms, Inc. | Generating multi-perspective responses by assistant systems |
US11245646B1 (en) * | 2018-04-20 | 2022-02-08 | Facebook, Inc. | Predictive injection of conversation fillers for assistant systems |
US11368420B1 (en) | 2018-04-20 | 2022-06-21 | Facebook Technologies, Llc. | Dialog state tracking for assistant systems |
US11231946B2 (en) | 2018-04-20 | 2022-01-25 | Facebook Technologies, Llc | Personalized gesture recognition for user interaction with assistant systems |
US11429649B2 (en) | 2018-04-20 | 2022-08-30 | Meta Platforms, Inc. | Assisting users with efficient information sharing among social connections |
US20210224346A1 (en) | 2018-04-20 | 2021-07-22 | Facebook, Inc. | Engaging Users by Personalized Composing-Content Recommendation |
US11544305B2 (en) | 2018-04-20 | 2023-01-03 | Meta Platforms, Inc. | Intent identification for agent matching by assistant systems |
US11676220B2 (en) | 2018-04-20 | 2023-06-13 | Meta Platforms, Inc. | Processing multimodal user input for assistant systems |
US11249774B2 (en) | 2018-04-20 | 2022-02-15 | Facebook, Inc. | Realtime bandwidth-based communication for assistant systems |
US11688159B2 (en) | 2018-04-20 | 2023-06-27 | Meta Platforms, Inc. | Engaging users by personalized composing-content recommendation |
US11704900B2 (en) * | 2018-04-20 | 2023-07-18 | Meta Platforms, Inc. | Predictive injection of conversation fillers for assistant systems |
US11704899B2 (en) | 2018-04-20 | 2023-07-18 | Meta Platforms, Inc. | Resolving entities from multiple data sources for assistant systems |
US11715289B2 (en) | 2018-04-20 | 2023-08-01 | Meta Platforms, Inc. | Generating multi-perspective responses by assistant systems |
US11715042B1 (en) | 2018-04-20 | 2023-08-01 | Meta Platforms Technologies, Llc | Interpretability of deep reinforcement learning models in assistant systems |
US11721093B2 (en) | 2018-04-20 | 2023-08-08 | Meta Platforms, Inc. | Content summarization for assistant systems |
US11727677B2 (en) | 2018-04-20 | 2023-08-15 | Meta Platforms Technologies, Llc | Personalized gesture recognition for user interaction with assistant systems |
US11887359B2 (en) | 2018-04-20 | 2024-01-30 | Meta Platforms, Inc. | Content suggestions for content digests for assistant systems |
US11886473B2 (en) | 2018-04-20 | 2024-01-30 | Meta Platforms, Inc. | Intent identification for agent matching by assistant systems |
US11908181B2 (en) | 2018-04-20 | 2024-02-20 | Meta Platforms, Inc. | Generating multi-perspective responses by assistant systems |
US11908179B2 (en) | 2018-04-20 | 2024-02-20 | Meta Platforms, Inc. | Suggestions for fallback social contacts for assistant systems |
Also Published As
Publication number | Publication date |
---|---|
EP2047389A1 (en) | 2009-04-15 |
BRPI0714327A2 (en) | 2013-04-24 |
TW200817943A (en) | 2008-04-16 |
NO20085388L (en) | 2009-02-25 |
US7849079B2 (en) | 2010-12-07 |
AU2007281645B2 (en) | 2011-09-29 |
EP2047389A4 (en) | 2011-06-22 |
US20080027921A1 (en) | 2008-01-31 |
RU2435213C2 (en) | 2011-11-27 |
RU2009103144A (en) | 2010-08-10 |
CN101496009A (en) | 2009-07-29 |
JP2009545810A (en) | 2009-12-24 |
AU2007281645A1 (en) | 2008-02-07 |
MX2009000591A (en) | 2009-01-27 |
WO2008016498A1 (en) | 2008-02-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US7577718B2 (en) | Adaptive dissemination of personalized and contextually relevant information | |
US7849079B2 (en) | Temporal ranking of search results | |
US7685199B2 (en) | Presenting information related to topics extracted from event classes | |
US8661031B2 (en) | Method and apparatus for determining the significance and relevance of a web page, or a portion thereof | |
US8060513B2 (en) | Information processing with integrated semantic contexts | |
US8631001B2 (en) | Systems and methods for weighting a search query result | |
US11360986B2 (en) | Enterprise search | |
US11625426B2 (en) | Incorporating opinion information with semantic graph data | |
US10552429B2 (en) | Discovery of data assets using metadata | |
US20100005087A1 (en) | Facilitating collaborative searching using semantic contexts associated with information | |
US8478756B2 (en) | Contextual document attribute values | |
EP2013784A2 (en) | Method and system for providing focused search results | |
MXPA05001073A (en) | System and method for a unified and blended search. | |
US20180349500A1 (en) | Search engine results for low-frequency queries | |
US9152698B1 (en) | Substitute term identification based on over-represented terms identification | |
US20140095465A1 (en) | Method and apparatus for determining rank of web pages based upon past content portion selections | |
AU2011253732B2 (en) | Temporal ranking of search results | |
US20140149378A1 (en) | Method and apparatus for determining rank of web pages based upon past content portion selections | |
JP2011018152A (en) | Information presentation device, information presentation method, and program | |
KR101449994B1 (en) | Method for providing retrieval service according to user preference | |
US20140280046A1 (en) | Searching using social filters as operators | |
Bozzon et al. | A characterization of the layout definition problem for web search results | |
Renda | A PERSONALIZED INFORMATION SEARCH ASSISTANT | |
Renda | Personalized Information Search and Retrieval through a Desktop Application |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
STCB | Information on status: application discontinuation |
Free format text: ABANDONED -- FAILURE TO RESPOND TO AN OFFICE ACTION |
|
AS | Assignment |
Owner name: MICROSOFT TECHNOLOGY LICENSING, LLC, WASHINGTON Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:MICROSOFT CORPORATION;REEL/FRAME:034544/0001 Effective date: 20141014 |