WO2006116194A2

WO2006116194A2 - Optimal word and phrase placement method and system for governing search results ranking

Info

Publication number: WO2006116194A2
Application number: PCT/US2006/015260
Authority: WO
Inventors: Timothy James Stanley
Original assignee: Timothy James Stanley
Priority date: 2005-04-21
Filing date: 2006-04-20
Publication date: 2006-11-02
Also published as: WO2006116194A3

Abstract

A method and system for optimizing search results ranking through optimal word and phrase placement is disclosed. The inventive method provides for efficient retrieval of relevant responses to queries mediated by search engines. Statistical weighting of word and phrase placement is used to satisfy ranking goals. Alternate embodiments include optimization for non-word searcher input and results such as audio or audio- visual combinations.

Description

TITLE

Optimal Word and Phrase Placement Method and System for Governing Search Results Ranking

RELATED APPLICATIONS

This application claims priority from provisional application no. 60/673,591, filed April 21, 2005, of the same title and same inventor, the entirety of which is incorporated by reference as if fully set forth herein.

GOVERNMENT FUNDING Not Applicable.

BACKGROUND

The internet functions for many as a de facto database from which information may be gleaned by means of search engines. Information providers desire to position information to be readily retrieved in response to information seekers, hi the case of commercial enterprises, achieving a high ranking in the search results provides increased opportunity to connect with the information seeker. What is needed is a means to arrange information accessible for search in order to optimize search results ranking so that relevant content shows up well for relevant search terms.

BRIEF SUMMARY OF THE INVENTION

The invention provides a method for optimizing search results ranking through optimal word and phrase placement within document text so that relevant content i shows up well for relevant search terms. The inventive method provides for efficient retrieval of relevant responses to queries mediated by search engines. Briefly, the inventive method includes the steps of:

Selecting a desired ranking criteria; obtaining keywords from a source; generating a word set from the keywords; evaluating the weight of each word set member; identifying placement options in the searchable web text; determining the numerical value of one or more word set members in one or more placement options; composing at least one optimized phrase from one or more word set members; inserting the optimized phrase in placement option and testing the ranking achieved when running a search through a search engine where the searchable text has inserted at least one optimized phrase in placement option; evaluating the ranking results form the test search to determine if the ranking criteria is satisfied; if yes, then no more steps are needed; if no, then more optimized phrases are composed, and the process steps repeated unit the ranking criteria is satisfied, or until it is determined that the ranking criteria cannot be satisfied.

A system according to the invention may be thought of as a communication system between an Author or Publisher of Media and a User or Information Seeker. The media is optimized: enhanced by means of the statistical placement of searchable words and phrases so that information/media content relevant to search queries appears strongly associated as indicated by result ranking. A User or Information Seeker instigating such an information search does so by means of a communication device, where the device is capable of a) composing and transmitting search queries mediated by search engine(s), and b) communicating search results to User such that in results, the Media is ranked in the search results communicated to User in an order strongly influenced by the optimization selected by the Publisher or Author. BRIEF DESCRIPTION OF THE DRAWINGS

Fig 1 is a flow chart depicting steps of the preferred embodiment of the inventive method.

Fig 2 schematically depicts the preferred embodiment of the inventive system.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT

The invention provides a method for optimizing search results ranking through optimal word (and phrase) placement within searchable document text. "Optimization" as used herein refers to the strong association between search terms and searchable media content terms such that the association results in ranking position in a query result. The inventive method provides for designing searchable media in order to ensure a ranking among any responses to statistically likely queries mediated by search engines.

More specifically, the invention provides a method of information placement within a website or other searchable construct where, owing to the statistical search strategies of search engine users, and taking into account the function of the search engines themselves, the relative ranking of the information placement as a response to a search query can be selected. The inventive method optimizes the results ranking and creates conditions favorable to a particular website or construct appearing in a preferred position ranking in search results obtained in response to queries mediated by database

(internet) search engines.

While the most common goal for commercial applications is to achieve a ranking high - i.e., to be among the first displayed to the Querying entity, internet User or information seeker - it is conceivable that the desired ranking could be low, or perhaps even a relative or proximity ranking is desired (always within x of y, where y is a collaborative researcher's media citation, or a competitor's website, for example). The inventive system includes the author, publisher, or information provider (perhaps a business or service), a means of placing searchable text in a position to be searched by search engines, a means for search queries to be posed to the search engines, and, in the preferred embodiment, a means for the User or Information Seeker to query.

Many words as used herein are used as commonly understood in current communication, m the interest of clarity, and not in limitation of the invention, the following words are used at least as set forth below, and not in a scope intended to curtail a full and complete understanding of the word or phrase.

Definitions

keyword: a word from data sources empirically determined to be relevant to actual queries posed in search engine mediated searching.

numerical value: a value combining word set member weight and placement values.

offering: the good or service or information for which some search result ranking is desired

page: means "web page" or some unit of search engine searchable media.

optimized phrase: one or more Word Set member words selected as a result of applying predetermined rules; such rules may include, for example, the number of words and/or the order of the words by word group.

placement option: position for insertion of at least one optimized phrase within a website, as defined herein.

placement value: value of a placement option or combination of placement options. preference value: a means of describing the "worth" of a search term (where a term could be a symbol, word, phase or any searchable expression);

refinement: dynamic adjustment of insertions of optimized phrase in placement options based on calibration of change in preference values over time.

website: the largest unit of searchable text for the purpose of determining ranking; a website may be more than one site and may include multiple sites connected by linking.

weight: the relative preference in search usage of a word or words within word groups; source data may be used in assigning weight; for example, if in a word group containing term X and term Y, term X appears in search queries more often than term Y, then the weight of term X is greater than the weight of term Y. "Weight" also includes data regarding the relation of the term to the sale of goods or services, and other factors.

word group: word set member(s) that are categorizable along some identifiable axis. For example, "locality" may be an identifiable axis, and a word group along that axis could contain word set member words pertinent to "locality".

word set: the set of keywords selected to be used for optimization. A word set member is a word that is a member of the word set.

Brief Recitation of the Method.

Fig 1 is a flow chart showing steps of the inventive method. First, there is the step of selecting a desired ranking criteria 10, followed by the steps of : obtaining keywords from a source 20; generating a word set from the keywords 30; evaluating the weight of each word set member 40; identifying placement options in the searchable website text 50; determining the numerical value of one or more word set members in one or more placement options 60 ; composing at least one optimized phrase 70 from one or more word set members; inserting the optimized phrase in placement option 80 [OPIPO].

After the insertion of the OPlPO, the resultant ranking obtained in response to a query is determined by a test query. Thus, testing the ranking 90 is achieved by running a search through a search engine where the searchable text has inserted at least one optimized phrase in placement option. After testing, the step performed is that of evaluating the ranking results from the test search to determine if the ranking criteria is satisfied 100. If the ranking criteria is satisfied, then the method is done [if yes 110, then no more steps are needed]. However, if the ranking is not satisfied, then repeat the steps, returning to the step of composing more optimized phrases 70 and repeat the process steps until the ranking criteria is satisfied - or until it has been determined that the result criteria cannot be satisfied.

Elaboration on the method.

The Author or Publisher - i.e., any entity causing the publishing of searchable media (media) - or one assigned with optimization ( hereafter, inclusively referred to as Publisher) must , according to the method taught herein, first "select ranking criteria" 10. Most common where the media is a tool for making a sale of a good or service, a ranking among the top search results is desired. In this light, for the purposes of discussion herein, let us use a hypothetical example: the Publisher is a business entity which selects a high ranking - namely, an appearance of results germane to the business within the top three response to a search query.

To obtain keywords 20, data sources (often marketing databases, industry databases, or other data sources) are checked and used to generate a list of terms actually used by those performing searches. The list of keywords will include those demonstrably used by Users or information seekers in search queries for the good or service being offered by Publisher.

A "User", or information seeker, refers to any entity composing and sending queries from a communication device enabling two-way communication: a transmission of queries that will be mediated by search engines, and a response capability, i.e., communicating search results to User. The Publisher has some customer profile in mind, as well as a description of the goods or services being offered. From a list of words potentially describing the offering, only those words that are also Keywords - words actually searched on - will generate the word set 30.

Members of the word set may be separable into " word groups". A common set of word groups are: locality, specific attributes, and general attributes.

Ex. For an example of a pizza delivery service located in Menlo Park, a list of keywords is: Redwood City, Menlo Park, mushroom, pepperoni, anchovy, pizza, spaghetti, calzone.

Accordingly, the word groups are:

Locality: Specific attributes General attributes Menlo Park mushroom pizza

Redwood City anchovy calzone

I. Locality In choosing keywords that relate to locality, the statistical data sources would be used to determine: a) how far (if at all) consumer travels to obtain the Offering; and b) how (what words are used) by the relevant consumer in searching for an Offering. Once it has been determined how far a consumer travels to obtain a product or service, then c) a list of locality keywords are obtained. For example, one may obtain a list of locality key words by: neighborhood, city, county, metropolitan area state d) using the data from a and b above, determine the level of search: i.e., the level that corresponds to the maximum distance a consumer seeks the good or service, d.l group sub areas that are a part of a larger area if both are typically searched e.g. if one searches for lawyers in Atlanta and in Georgia, then the words/ (partial phrase) "Atlanta, Fulton County, Georgia" could be used.

II. Specific Attributes Most service or products offerings have specific terms germane to the product or service, and these specific terms are included in the keyword selection. For the pizza example above, the specific attributes were mushroom and anchovy - the level of particular ingredients.

III. General Attributes

General attributes generate keywords that are the most searched on for the industry that relates to service or product being offered, with an emphasis on those that most closely match search terms that will lead to a sale or other desired transaction. For example, for legal services, the words Lawyer, Attorney, Lawyers, Attorneys, Law Firm, are more likely to lead to the hiring of a lawyer than the terms "Law" or

"Legal".

Next step: Evaluate Weight of each Keyword 40

Again, statistical data is used to determine the relative search frequency as between and among keywords, and a relative weighting can be assigned. Referring again to the pizza example, the word "pizza" might be included in more searches than "calzone" therefore, a relative weighting of pizza as greater than calzone can be assigned.

Such values can be assigned on a relative basis in each word group. This keyword weighting will interact with placement option evaluations to render a numerical value of any keyword in any placement option. Referring again to Fig 1, the step of "identify placement options 50" is performed after word set generation and weighting. The prevalence of placement options is related to the design and intended scope of the searchable media. If the media is a single website, for example, each web page can have multiple placements options (potential locations for inserting the statistically selected words or phrases as described herein). For example, there may be, say, 8 potential locations for phrase insertion on a web page: 2 placements may exist in the <title> tag, 1 placement in the <hl> tag, 1 placement in the <h2> tag, 4 placements in the body text.

The next step is to "determine numerical value of a word set member in placement option 60". hi this step, a numerical value is assigned for a word set member in each of the identified placement options single or in combination with other placement. Each placement option, or group of placement options, is given a numerical value relative to other placement options. The numerical value represents how strong that placement option is with regard to producing the desired (i.e., higher) results in search rankings. The word set member "pizza", in our pizza delivery example, may have a greater numerical value in the placement option in one of the <title> tag placements than in the body text placement. The numerical value of a placement option is a function of several aspects of weighting, including the position of the placement option on the website, the total number of placement options within the website, and the proximity to other word set members.

Next step: Compose optimized phrase 70

Once the placement options have been identified and assigned numerical values, the next step is to compose optimized phrases 70. An optimized phrase is a phrase comprising one or more word set members, alone or in combination, that will be inserted into placement option 80 and subsequently tested to determine if the ranking criteria is satisfied with that inserted OPIPO [optimized phrase in placement option].

An optimized phrase may be composed from the word set and, if there are more than one word groups in the word set, then a number of word groups. The optimized phrases are made by combining the selected word set members in an order that makes sense when reading; alternatively, the word set members may be combined so as to form phrases that are demonstrably υ^ed for searching. For many professions, an exemplar phrasing is as follows:

"Locality Key Word

Service Specific Keyword General Keyword" e.g., San Francisco

Medical Malpractice Lawyer

After the composition of optimized phrases, the next step is to insert the optimized phrases in placement options [OPIPO] 80. Certain rules may exist for the inserting of OPIPO. In the examples discussed herein, word group 1 is locality. Locality keywords are chosen based on statistics related to the entire website. In certain circumstances, there may be an event or a characteristic that supplies a specific keyword.

For example, there may be a specific locality keyword, such as "Macon County" placed in for workers accidents - if such accidents happened in a particular county where a manufacturer had a chicken processing plant and the goal was to reach workers in that chicken processing plant.

Alternatively, however, if the placement was for trucking accidents, then the more useful steps would be: obtaining a list of all the localities along the major freeway, and choosing a locality for any particular page statistically from the list of localities relevant based on initial conditions and selection criteria.

In determining placement options for optimized phrases with specific keywords, such as those related to a professional specialty or specific expertise, placement options are favored where the content on a web page or the website as a whole bear some relation to the specific keywords. Similarly, in selecting placement options for optimized phrases containing general keywords, it is useful to consider pages whose content include specific keywords or where a placement option exists for an optimized phrase containing specific (attribute) keywords. Next step: Test Inserted OPIPO 90

Once an optimized phrase in placement option [OPIPO] has been inserted, the method provides for a test search, sending a query via one or more search engines. The results of the query, that is to say, the ranking, are then examined to determine if the ranking criteria has been satisfied 100. If the ranking criteria has been satisfied, then the media has been optimized ( as defined herein) and the method is done 110. If the ranking criteria is not satisfied, then the process returns to the step of composing optimized phrases 70, and the steps of inserting and testing OPlPO until the ranking criteria is satisfied, or until it is determined that the ranking criteria cannot be satisfied.

Returning to our pizza example, the optimized phrase " Redwood City mushroom pizza" is inserted into a placement option in the website. Upon testing search results ranking with at least one query, the pizza store in question is displayed as the second store listed. This satisfies the criteria of achieving a ranking

(listing) in the top three.

The method further provides for periodic refinement: dynamic adjustment of insertions of optimized phrase in placement options based on calibration of change in preference values over time. The method provides for the capability of automatic re-optimization by a test of the search engine, or a repetition of some or all of the method steps to achieve desired ranking in search results.

Preference values can change over time owing to many factors. Some common sources of change to the searchable media content or its environment include: a. new pages added or deleted; b. changes in the rules by the search engines (this may lead to new placement options, the removal of certain placement options, a change in how placement options are grouped or ungrouped, a change a placement option's weight on a page); c. changes in the words or emphasis the publisher/author/information provider wants to focus on (for example: change in localities, changes in locality population, changes in practice, changes in terms and phrases for services or products, changes in prices for services); d. changes in the number of pages indexed by the search engines for the words and phrases relative to other pages and difficulty of competing for the words and phrases; e. analysis of click-throughs for words and phrases, and relation to "conversion" (what "converts" a visitor into a buyer, or what is/are the conversion paths characterizing a website visitor as the visitor is becoming a purchaser)

The inventive method provides for the refinement, the "re-optimization', of media to account for these and other such changes. Such refinement may be done automatically, periodically, at any pre-determined interval, or triggered by detected changes in the media or search engine behavior.

All steps in the inventive method may be implemented iteratively, through human acquisition and interpretation of search performance and ranking results. Alternatively, all or portions of the process may be automated by means of appropriate codification of the process steps.

THE INVENTIVE SYSTEM

Referring now to Fig 2, a system 200 according to the invention may be thought of as a communication system between a publisher or media source author 208 of searchable, published media 210 and a User, or information seeker, using a query composition and response communication device 240.

The searchable published media 210 is enhanced by means of the statistical placement of searchable words and phrases, i.e. statistical optimization 220, and is thereafter optimized searchable media 230. A User (not depicted) queries by means of a communication device 240 , wherein said device is capable of a) composing and transmitting search queries 242 mediated by search engine(s) 250, and b) communicating search results 248 to User such that the searchable published media 230 is ranked in the search results communicated to User in an order influenced by the optimization imposed by the publisher 208.

Statistical optimization 220 includes both initial test of the optimization and refinement and update, all of which require running test queries 219 via the search engine 250 under characterization. The ranking response 217 of the search engine to a optimization test query 219 determines whether the ranking criteria has, has not been - or in some instances, cannot be - satisfied.

User initiated query 242 mediated by a search engine 250 engages several aspects. The search engine has, at some point in time, searched all searchable media, and such search includes as search of the optimized searchable media 244; the search engine results of a search of the optimized searchable media 246, are incorporated into the search engine's content database 252 and, in turn, in the index 254. Thus, any response form the search engine 250 to a user query 242 necessarily engages the index

254 and reflects the state of the content database 252.

It bears repeating that although the examples used herein relates to commercial goals, the invention is applicable to areas including education or research, as well as venues or databases that may be different from the internet as it currently exists. While the output in the instant invention relates to visual search results, the optimization as taught by the invention may be applied as well to audio output or visual output not primarily word based or any combination.

The embodiments and examples set forth herein are merely illustrative of the principles and applications of the present invention. Numerous modifications may be made to the illustrative embodiments and other arrangements may be devised within the scope of the present invention as taught by the specification, the drawings, and any appended claims.

Claims

WE CLAM:

1. A method for influencing search results ranking ("ranking") of published searchable media in response to queries mediated by search engines, said method comprising the steps of: selecting a desired ranking criteria; obtaining keywords from a source; generating a word set from the keywords; evaluating the weight of each word set member; identifying placement options in the searchable web text; determining the numerical value of one or more word set members in one or more placement options; composing at least one optimized phrase from one or more word set members; inserting the optimized phrase in placement option and testing the ranking achieved when running a search through a search engine where the searchable text has inserted at least one optimized phrase in placement option; evaluating the ranking results from the test search to determine if the ranking criteria is satisfied : i. if yes, then no more steps are needed; ii. if no, then return to the step of composing more optimized phrases and repeat the process steps until the ranking criteria is satisfied or has been shown to be unsatisfiable.

2. A system enabling search engine mediated communication between a media publisher

(information provider) and a User (information seeker), said system comprising: a means for publishing media, wherein the media is: a) searchable by means of at least one search engine, and b) optimized: i.e., enhanced by the statistical placement of searchable words and phrases a communication device capable of: a) composing and transmitting search queries mediated by search engine(s), and b) communicating search results to User such that the media is ranked in the search results communicated to User in an order influenced by the media publisher.