WO1997038376A2 - A system, software and method for locating information in a collection of text-based information sources - Google Patents
A system, software and method for locating information in a collection of text-based information sources Download PDFInfo
- Publication number
- WO1997038376A2 WO1997038376A2 PCT/IB1997/000748 IB9700748W WO9738376A2 WO 1997038376 A2 WO1997038376 A2 WO 1997038376A2 IB 9700748 W IB9700748 W IB 9700748W WO 9738376 A2 WO9738376 A2 WO 9738376A2
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- linguistic
- terms
- text
- knowledge base
- language
- Prior art date
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/3332—Query translation
- G06F16/3338—Query expansion
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/3332—Query translation
- G06F16/3334—Selection or weighting of terms from queries, including natural language queries
Definitions
- the present invention relates generally to the field of information retrieval. More particularly, the present invention relates to information management systems and computational linguistic systems for finding information related to a user-input query, in a collection of text- based information sources.
- data is stored in a strictly structured environment.
- Such systems may be based upon tables of records or spreadsheet models, for example.
- Such systems may be flat or may be relational with respect to how records in the database are associated with each other.
- conventional database management systems generally require structured records in which one or more fields may be searchable, i.e. are key fields.
- key fields use terms, e.g. numbering systems, labels, etc., in a consistent manner which facilitates searching with known query values, i.e. combinations of numbers, labels, etc.
- full-text searching has developed.
- Full-text searching of a collection of text-based information sources permits a user to write a query containing terms known to be used in relevant documents.
- the collection of documents is first fully indexed and the words of the documents in the index are compared with the query terms.
- an exact match between a query term and an index entry must be found in order to identify a relevant document. Spelling errors, word variants, etc. will tend to prevent finding all relevant documents.
- a technique called wild- carding may be used to partially alleviate this problem, but many irrelevant documents, referred to as "noise,” often turn up when wild-carding is used.
- Pinkas G., Natural Language Full-Text Retrieval System, Master's Thesis, University of Jerusalem, 1985, discloses a system which automatically expands a user's query to include additional relevant terms in a manner more noise-free than simple wild-carding.
- the Pinkas system (1) receives a user query composed of query-words and boolean operators; (2) expands the query linguistically, i.e. by referring to a pre-processed database of morpholo_gical and phonetic information; (3) expands the query associatively, i.e.
- query expansion depended upon a set of linguistic rules which were developed by an expert in the language to be processed.
- the set of linguistic rules was both extensive and relatively inflexible, since as many characteristics of the input language as possible had to be accounted for before processing any text-based information sources.
- Development of the linguistic rules for each language to be processed was a very labor-intensive and time- consuming task.
- aspects of the present invention solving the problems of the prior art include at least a system, software and a method for processing information contained in a collection of text-based information sources.
- the system may include a computer or data processor and software structured as one or more software modules, units or functions which when executed in a specified order by the computer or data processor perform the desired information processing task.
- One or more software modules, units or functions may be made available in conventional manner as either compile-time or run-time library entries which may be referred to by a software program which is written in a manner to be aware of such a library.
- the present invention further provides a method to process query-concepts and transform them to an expanded/improved query using an expansion matrix providing great flexibility, high accuracy and low noise output.
- a text-based information processing system comprising an automatic linguistic knowledge base generator having an input receiving a collection of text-based information sources and which produces a linguistic knowledge base; an index generator having inputs receiving a collection of text-based information sources and the linguistic knowledge base and which produces an index of the received text-based information and further which updates the linguistic knowledge base to reflect the inputs to the index generator and maintain correlation between the index and the linguistic knowledge base; a query processor having inputs receiving a query composed by an operator, the linguistic knowledge base, the index and a thesaurus and which produces a list of locations in the collection of text-based information sources relevant to the query.
- the text- based information processing system may be subject to numerous modifications and variations.
- the automatic linguistic knowledge base generator, the automatic index generator and the query processor may be embodied in various ways.
- an automatic linguistic knowledge base generator may comprise a parser, receiving an input stream of terms and producing individual terms; a language recognizer connected to receive the individual terms from the parser and which produces an output indicative of a language to which each individual term belongs; a normalizer connected to receive the individual terms and further to receive linguistic rules for the language indicated by the output of the language recognizer and producing normalized terms; and a linguistic expander connected to receive the legal individual terms and producing entries stored in the linguistic knowledge base.
- an automatic indexer may comprise a parser, receiving an input stream of terms and producing individual terms; a language recognizer connected to receive the individual terms from the parser and which produces an output indicative of a language to which each individual term belongs; a normalizer connected to receive the individual terms and further to receive linguistic rules for the language indicated by the output of the language recognizer and producing normalized terms; and an index entry generator connected to receive the legal individual terms and producing entries stored in the index when the terms have not previously been indexed and modifying an existing index entry when the terms have previously been indexed.
- an expansion unit for expanding terms in a language may comprise an associative expander having an input receiving a term and having an output representing the term and at least one associated term found by the associated expander making reference to a thesaurus; and a linguistic expander having an input connected to the output of the associative expander and having an output representing the input of the linguistic expander and at least one term linguistically related to the input of the linguistic expander and found by reference to a linguistic knowledge base for the language.
- the normalizers recited above may be constructed of two units.
- the first normalizer unit may be connected to receive the individual terms and the linguistic rules and producing terms from which illegal characters have been removed; and the second normalizer unit may then be connected to receive the terms from which illegal characters have been removed and the linguistic rules and which produces normalized terms including word stems found by applying the linguistic rules to the terms from which illegal characters have been removed.
- Fig. 1 is a schematic block diagram of a computer or data processing system on which the present invention may be practiced;
- Fig. 2 is a schematic block diagram of the memory of Fig. 1;
- Fig. 3 is a flow chart of automatic linguistic knowledge base generation;
- Fig. 4 is a flow chart of automatic index generation;
- Fig. 5 is a flow chart of query expansion
- Fig. 6 is a flow chart of an information retrieval system including the features illustrated in Figs. 3-5.
- a "language” is considered to be any organized system of tokens, which have symbolic meaning.
- the tokens are referred to hereinafter as "words” or “terms,” since the most common types of languages dealt with by text- based information systems are natural human languages composed of words or combinations of words, i.e. terms, which are understood to have specific meanings by humans.
- words and “terms” are intended to encompass word phrases in those instances where a word phrase may in fact be a token having a meaning independent of its sub-units, keywords in those instances where a word or word-phrase has a specific context/importance, and artificial words such as acronyms and short-cuts.
- a “word base” is the base portion of a word which remains after removing all prefixes and suffixes of the word which modify the meaning or part of speech of the word root appropriately for the context in which the word may be used.
- thesaurus refers to a database of terms, words and/or word bases, in which each term, word or word base is associated in the database with other terms, words and word bases having a defined relationship such as mo ⁇ hological proximity, phonetic similarity, similar meaning (synonyms), nearly opposite meaning (antonyms), broader meaning, narrower meaning, related term in a specific context, etc.
- the database may be navigated or searched on the basis of the terms, words and/or word bases stored therein.
- Languages considered here have known linguistic rules for the mo ⁇ hological and phonetic variations which words may undergo.
- the mo ⁇ hological rules of a language may define how a plural is formed from a singular noun, by changing the shape of the word, i.e. adding a final "s" in English, while the phonetic rules may represent the common variations in spelling resulting from user spelling errors.
- a table, file or database may be used by a software program to hold a list of such linguistic rules.
- languages also include words which do not follow the linguistic rules of the language.
- the English language mo ⁇ hological rule for generating the past tense of a verb does not apply to the verb "to go,” which becomes "went” rather than the nonsensical "goed.” Therefore, exceptions to the rules may be held by a software program in a table of exceptions, so that words which do not obey the rules may be handled as accurately as words which do obey the rules.
- a "linguistic knowledge base" is developed by applying the linguistic rules and the one or more tables of exceptions or irregular forms to a large body of textual information to produce an efficient, adaptable and flexible representation of the variations of word bases which produce meaning, particularly in natural languages, but also in language generally.
- a computer system or data processing system generally includes a processor 101, a memory 103, one or more input devices 105, and one or more output devices 107, all interconnected through an interconnection mechanism 109.
- processor 101 a processor 101
- memory 103 a memory 103
- input devices 105 a processor
- output devices 107 all interconnected through an interconnection mechanism 109.
- an interconnection mechanism 109 Many variations of this basic plan are possible. For example, viable systems may lack input devices 105 and output devices 107, communicating entirely through interactions with the memory 103 by external devices (not shown). Also, distributed computer systems and data processing systems are contemplated as falling within this basic plan.
- the interconnection mechanism 109 may be an internal system bus of a personal computer or may be the Internet, through which a processor 101 interacts with a database stored on a remote memory 103.
- Memory 103 may be classified into two categories useful to this discussion, long term memory (also called non-volatile memory), and short term memory (also called volatile memory). These two types of memory are often both used in computer systems and data processing systems, as shown in Fig. 2.
- Volatile memory 201 such as integrated circuit random access memory (RAM) is often used in close physical proximity to the processor 101 because the technologies in which such volatile memory 201 is most readily realized produce fast access times, such as are desirable to support fast processors 101.
- RAM integrated circuit random access memory
- Non-volatile memory 203 is often used to store massive quantities of data for longer periods of time because it can be more cheaply constructed than volatile memory of a similar capacity.
- Non-volatile memory 203 is often implemented as magnetic or optical disk or tape storage units, which provide a further advantage of data and software program interchange between different computer or data processing systems.
- non-volatile memory 203 may be a software product disk on which are recorded signals representing instructions, which when executed by a processor 101 cause the computer or data processing system to perform a special pu ⁇ ose function.
- Software embodying aspects of the present invention may be recorded on such a non-volatile memory 203 for distribution by a manufacturer, for archival pu ⁇ oses, for access through a volatile memory 201 by a processor 101. etc.
- a system for searching through and locating information in a collection of text-based information sources may be constructed.
- a linguistic knowledge base is first generated. Then, the collection of text-based information sources is indexed. A user next inputs a query defining the information sought. The query is expanded according to selected associative and linguistic rules, using a thesaurus and the linguistic knowledge base. Finally, information is identified which matches the various expanded query terms.
- the thesaurus, linguistic knowledge base and index may be stored in one or more computer files to which the system has access through memory 103.
- the aspects of the invention connected with automatic generation of the linguistic knowledge base, automatic generation of the index and query expansion are next described in detail. I. Automatic Generation of the Linguistic Knowledge Base
- Fig. 3 software as shown in Fig. 3 is provided which when executed by a suitable data processing system will automatically generate the linguistic knowledge base from an input body of text based information sources.
- a collection of English language documents may be processed to generate an English language linguistic knowledge base.
- a small set of linguistic rules 301 including a list of exceptions 302 to the linguistic rules for a language, e.g. English, is first generated by statistical analysis of a large body of text based information.
- This small set of rules includes:
- a word normalization table specifying legal characters in the language, i.e. the alphabet of the language, and legal character positions in the language, e.g. special rules concerning characters which can only appear at specific locations within a word;
- This set of rules 301 is then used to analyze a body of text based information sources 303, to generate a linguistic knowledge base 305 specifically adapted from the body of text based information sources 303.
- the body of sources 303 may be selected to be sources from a particular field of endeavor in which future queries are expected to be made, for example. This will result in a linguistic knowledge base better able to cope with the specifics of that particular field of endeavor.
- the body of sources 303 from which the linguistic knowledge base 305 is derived may not be the same body of sources which is ultimately to be searched. However, automatically generating the linguistic knowledge base 305 from the body of sources to be searched has the advantage that the linguistic knowledge base 305 so produced is particularly well adapted to the body of sources to be searched.
- Automatic generation of the linguistic knowledge base 305 proceeds as follows.
- the body of text based information sources 303 forms an input stream of text 304 to the system.
- This input stream 304 is first parsed into words and terms 307 in accordance with either fixed word recognition rules or word recognition rules specific to one or more languages.
- the language of each of the words parsed from the input stream is then recognized 309. Once the language of a word has been recognized the word may be normalized 311 according to the linguistic rules 301 for the language. Irregular words may also be recognized at this point, since known irregular words are already in the list of irregular words 302 and hence need no further processing.
- the system may also identify as potential new irregular words, those words meeting some rule-based criteria.
- Regular words are linguistically expanded 313 before being added to the linguistic knowledge base 305 such that word bases are stored in the linguistic knowledge base 305 along with a list of related words from the body of sources 303. Linguistic expansion 313 is discussed in greater detail below.
- the step of parsing 307 the input stream 304 into sentences and words takes place according to the following pseudo-code:
- ⁇ for each word ⁇ if language not explicitly specified, identify word's language end if; normalize word; return word and word's coordinates; return rest of input stream; ⁇ next word; ⁇ next sentence.
- Normalization is performed as follows. Normalization identifies and removes garbage characters from the words of the input stream 304. For each character of an input word
- new keys are added to the linguistic knowledge base 305 by the following procedure. If language not explicitly specified identify the language of the input word; for each recognition type
- the mo ⁇ hological analyzer of the described embodiment functions in accordance with the following procedure.
- the mo ⁇ hological analyzer receives a list of valid prefixes and suffixes in the language identified for the input word. Start at end of word; strip next substring from end; for each substring of word ⁇ /* search for prefix*/ if substring is found to be a prefix in the identified language
- the phonetic analyzer converts each word into a phonological representation of the word on the basis of letter to sound rules. Words having similar or same phonological representations may be considered to be related by their phonetic mo ⁇ hology.
- a linguistic knowledge base 305 for the languages of the text in the body of sources 303 will have been automatically generated.
- new text based information sources are added to the system, they are also processed as above.
- new sources increase the knowledge and accuracy of the linguistic knowledge base 305 through the addition of new information to the linguistic knowledge base 305, as well as through the linguistic correction mechanism which corrects the contents of individual entries in the linguistic knowledge base 305 according to new information.
- the learning procedure which embodies the linguistic correction mechanism is as follows. Open a new table entry for a new correct key; get the list of words in the body of a previous key entry; for each word
- the affected word base and list of related words may be automatically, or at the direction of a human operator, reanalyzed and updated in accordance with the newly presented information and the above procedure.
- the system constantly learns about each language processed and updates the affected linguistic knowledge bases.
- the retrieval system In addition to the linguistic knowledge base 305, the retrieval system according to another aspect of the present invention shown in Fig. 4 automatically generates an index 401, whereby text based information may be found by reference to the index 401. Automatic generation of the index 401 is accompanied by updating of the linguistic knowledge base 305, so that the contents of the linguistic knowledge base 305 reflects the relevant terms contained in the body of text based information sources 303 and is thus correlated with the index 401.
- the index 401 simply relates words actually found in the body of text based information sources 303 to locations within the body of sources 303. It is preferred that the location be defined hierarchically. For example, the location may be represented hierarchically by a document number, section number, sentence number and position number. Other hierarchical location identification schemes may be used, as seen fit by those skilled in this art.
- the index 401 is assisted by the linguistic knowledge base 305.
- the index 401 includes only words and terms actually occurring in the text-based information sources 303.
- the linguistic knowledge base 305 relates word bases derived from the words actually occurring in the text-based information sources 303 to lists of related words. During retrieval, which is explained below, the system retrieves an entry from the linguistic knowledge base 305 which is then used to reference one or more index entries.
- Automatic generation of the index 401 proceeds as follows.
- the body of text based information sources 303 forms an input stream of text 304 to the indexing subsystem.
- This input stream 304 is first parsed into words and terms 307 in accordance with the word recognition rules.
- the language of each of the words parsed from the input stream is then recognized 309. Once the language of a word has been recognized 309 the word may be normalized 31 1 according to the linguistic rules 301 for the language.
- An index entry is then generated 403 for each new normalized word. If the normalized word already has an entry in the index 401 , then the location of the current occurrence of the word is added to the previous entry.
- the linguistic knowledge base 305 is continuously kept correlated with new and modified entries produced in index 401.
- Each normalized word is reduced to its word base 405 in accordance with the linguistic rules of the language of the word.
- the word base and related word is then added to the linguistic knowledge base file 407, if not already present.
- the user may also specify that related words include various types of expansions of the word bases. If expansions are included, expansion of the word base is performed before storing the word base and related words in the linguistic knowledge base file 305.
- the linguistic knowledge base 305 is correlated with the index 401 and reflects the relevant terms contained in the body of text based information sources 303.
- Query expansion is performed in accordance with a third aspect of the invention, shown in Fig. 5. Since a query may contain more than one word or term, word recognition is first performed as above. The words and terms identified by the word recognition task may further be normalized.
- Each recognized word in the query may then be expanded using a 2D expansion matrix.
- the 2D expansion matrix is one way of defining the expansion space in which an input word may be represented.
- the dimensions of this space are associative and linguistic.
- the associative dimension is based upon the meaning of words/word-bases in the language to be processed.
- the associative dimension is defined by one or more thesauri 501 relating words and terms to their synonyms, broader terms, narrower terms and other relations.
- Each thesaurus 501 includes a database of terms along with conceptually related terms.
- the thesaurus is searchable by term.
- each thesaurus entry contains an entry key which is a list of searchable terms.
- Each entry key has associated therewith one or more terms conceptually related to the entry key, such as synonyms, broader terms, narrower terms, associated terms, antonyms, etc. The inclusion of any one or more categories of association is optional.
- each entry term may optionally have associated therewith a conventional dictionary definition and usage guide, as well as a query string into which the entry key may be translated when required.
- the thesaurus is a list of entries, wherein each entry has a structure substantially as follows:
- KEYWORD (in the form of natural language phrase or term) Used as an entry key.
- SYNONYMS A list of keywords synonymous with the KEYWORD that comprise a concept or descriptor.
- the linguistic dimension of this expansion space is based upon the linguistic knowledge base 305 of the language to be processed.
- the linguistic knowledge base is built automatically from the actual co ⁇ ora of the text-based information sources, independent of manually crafted linguistic dictionaries, and not being restricted to "legal" or "proper” words.
- linguistic expansion grammars of mo ⁇ hology and phonetics are supported.
- the expansion task performs 2D expansion in substantially two main steps.
- First an associative expansion is performed at step 503, in which each input word of an input query 505 is expanded to a list of words 507 including words having defined relations to the input word.
- the associated words are found by making reference to the thesaurus 501.
- This expanded list of words 507 becomes the input on which linguistic expansion 509 is performed in both the mo ⁇ hological and phonetic dimensions, simultaneously.
- the mo ⁇ hological and phonetic expansion is controlled by making reference to the linguistic knowledge base 305.
- the linguistic expansion 509 may be controlled by expansion parameters 511 supplied by the user to include varying degrees of mo ⁇ hological expansion and phonetic expansion, ranging for both dimensions from no expansion in that dimension to full expansion in that dimension.
- expansion strategies for mo ⁇ hology and phonetics may be intelligently related.
- the relationships between the expansion dimensions are defined in the linguistic knowledge base 305 for the language.
- a rule for mo ⁇ hological expansion may define a mo ⁇ hological variation which changes depending upon the phonetic properties of the input word or the expanded result.
- less noise is generated in the expanded output because relating the mo ⁇ hological and phonetic dimensions as a single linguistic plane eliminates mo ⁇ hological variants which are phonetically unacceptable under the totality of the linguistic rules, and vice versa.
- a retrieval system may be constructed as shown in Fig. 6, which can perform efficient and accurate location of information within a collection of text-based information sources.
- a system is given access to one or more collections of text-based information sources 303a and 303b.
- At least one group of text-based information sources 303a is supplied to automatic linguistic knowledge base generating software 601, which generates the linguistic knowledge base 305 as described above.
- Text-based information sources 303b are provided to an indexing subsystem 603 which creates an index 401 of words in the text-based information sources 303b, in which each entry in the index 401 defines a relationship between a word and the location of the word in the collection, as described above.
- the index 401 be generated using normalized words in one or more languages for which the system has a thesaurus 501 and a linguistic knowledge base 305.
- the indexing subsystem 603 may include a module to recognize words having forms which conform to one of the languages supported by the system and may further include an appropriate normalizing module for each language supported by the system. Words are normalized in their language as discussed above, to reduce the number of anomalous entries appearing in the index 401.
- the system further receives a user query 505 in the form of one or more words expressive of the information sought by the user.
- the query words are expanded 605 using a 2D expansion matrix, as discussed above.
- the query is first associatively expanded to include words related to the original query words by reference to the thesaurus appropriate for the language of the query words.
- the associatively expanded query is then linguistically expanded in both the mo ⁇ hological and phonetic dimensions, simultaneously.
- the degree of expansion in each dimension is specified by the user, by parameters 511 supplied with the query.
- the degree of expansion may be specified by the user, for example, by attaching a checklist of expansion parameters 51 1 to each query term.
- the terms of the fully expanded query 607 are compared 609 with the entries in the index 401 to find relevant locations 611 within the collection of text-based information sources 303b.
- Relevant locations 611 in the collection of text-based information sources 303b do not necessarily contain any of the original query terms.
- the locations found 611 will contain one of the original query terms or a related term produced by the associative and linguistic expansion processes.
- the locations found 61 1 will not include many '"noise " locations because the linguistic expansion process is performed as described above in a manner in which the mo ⁇ hological and phonetic linguistic rules are applied simultaneously in a synergistic manner that avoids the problem of applying a mo ⁇ hological rule to generate a phonetically nonsensical result or vice versa.
- the text-based information sources may be text documents stored on a computer system.
- the indexing system may hierarchically refer to locations by document number, section number, sentence number and position with the sentence.
- freely formatted text documents may be processed by the above-described system. There is no need to structure the documents a particular way or to manually produce classifications or keywords, as done in some prior art systems, because the present system indexes words and manipulates queries according to the rules of the language in which the words occur. If it is desired that a phrase be treated as a single word or term in a particular language, then that phrase may be so defined as a conceptual entity in a thesaurus. In all other respects, the phrase so defined as a word is treated simply as a word in the language. However, it is unnecessary to declare a long list of accepted keywords because the process of indexing and query expansion generates accurate, relatively noise-free matches for user queries reasonably expressive of the information sought.
Abstract
Description
Claims
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
EP97925221A EP0934569A2 (en) | 1996-04-04 | 1997-04-04 | A system, software and method for locating information in a collection of text-based information sources |
JP9532080A JP2000507008A (en) | 1996-04-04 | 1997-04-04 | Systems, software and methods for locating information in a collection of text-based information sources |
Applications Claiming Priority (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US1481596P | 1996-04-04 | 1996-04-04 | |
US66047896A | 1996-06-07 | 1996-06-07 | |
US08/660,478 | 1996-06-07 | ||
US60/014,815 | 1996-06-07 |
Publications (2)
Publication Number | Publication Date |
---|---|
WO1997038376A2 true WO1997038376A2 (en) | 1997-10-16 |
WO1997038376A3 WO1997038376A3 (en) | 1997-12-04 |
Family
ID=26686566
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/IB1997/000748 WO1997038376A2 (en) | 1996-04-04 | 1997-04-04 | A system, software and method for locating information in a collection of text-based information sources |
Country Status (4)
Country | Link |
---|---|
EP (1) | EP0934569A2 (en) |
JP (1) | JP2000507008A (en) |
CA (1) | CA2250694A1 (en) |
WO (1) | WO1997038376A2 (en) |
Cited By (31)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP1104568A1 (en) * | 1998-07-15 | 2001-06-06 | Amazon.Com. | System and method for correcting spelling errors in search queries |
WO2001067280A2 (en) * | 2000-03-08 | 2001-09-13 | Accenture Llp | Method for a knowledge model |
WO2001098946A1 (en) * | 2000-06-22 | 2001-12-27 | Hapax Information Systems Ab | Method and system for information extraction |
GB2393541A (en) * | 2002-09-27 | 2004-03-31 | Hewlett Packard Development Co | Method for management of synonymic searching |
EP1411448A2 (en) * | 2002-10-17 | 2004-04-21 | Matsushita Electric Industrial Co., Ltd. | Data searching apparatus |
US6735560B1 (en) * | 2001-01-31 | 2004-05-11 | International Business Machines Corporation | Method of identifying members of classes in a natural language understanding system |
US6957205B1 (en) | 2000-03-08 | 2005-10-18 | Accenture Llp | Knowledge model-based indexing of information |
US6996774B2 (en) | 2002-02-12 | 2006-02-07 | Accenture Global Services Gmbh | Display of data element indicia based on data types |
GB2417115A (en) * | 2002-09-27 | 2006-02-15 | Hewlett Packard Development Co | Managing synonymic searching and ranking results |
US7072847B2 (en) * | 2000-08-25 | 2006-07-04 | Jonas Ulenas | Method and apparatus for obtaining consumer product preferences through product selection and evaluation |
US7089188B2 (en) * | 2002-03-27 | 2006-08-08 | Hewlett-Packard Development Company, L.P. | Method to expand inputs for word or document searching |
US7350138B1 (en) * | 2000-03-08 | 2008-03-25 | Accenture Llp | System, method and article of manufacture for a knowledge management tool proposal wizard |
EP2074505A4 (en) * | 2006-10-05 | 2010-01-13 | Splunk Inc | Time series search engine |
US7716201B2 (en) | 2006-08-10 | 2010-05-11 | Yahoo! Inc. | Method and apparatus for reconstructing a search query |
US7765176B2 (en) | 2006-11-13 | 2010-07-27 | Accenture Global Services Gmbh | Knowledge discovery system with user interactive analysis view for analyzing and generating relationships |
US7853556B2 (en) | 2003-09-12 | 2010-12-14 | Accenture Global Services Limited | Navigating a software project respository |
US7904411B2 (en) | 2005-02-04 | 2011-03-08 | Accenture Global Services Limited | Knowledge discovery tool relationship generation |
US7979314B2 (en) | 2001-08-23 | 2011-07-12 | Jonas Ulenas | Method and apparatus for obtaining consumer product preferences through interactive product selection and evaluation |
US8010581B2 (en) | 2005-02-04 | 2011-08-30 | Accenture Global Services Limited | Knowledge discovery tool navigation |
US8356036B2 (en) | 2005-02-04 | 2013-01-15 | Accenture Global Services | Knowledge discovery tool extraction and integration |
US8660977B2 (en) | 2005-02-04 | 2014-02-25 | Accenture Global Services Limited | Knowledge discovery tool relationship generation |
US10019496B2 (en) | 2013-04-30 | 2018-07-10 | Splunk Inc. | Processing of performance data and log data from an information technology environment by using diverse data stores |
US10225136B2 (en) | 2013-04-30 | 2019-03-05 | Splunk Inc. | Processing of log data and performance data obtained via an application programming interface (API) |
US10318541B2 (en) | 2013-04-30 | 2019-06-11 | Splunk Inc. | Correlating log data with performance measurements having a specified relationship to a threshold value |
US10346357B2 (en) | 2013-04-30 | 2019-07-09 | Splunk Inc. | Processing of performance data and structure data from an information technology environment |
US10353957B2 (en) | 2013-04-30 | 2019-07-16 | Splunk Inc. | Processing of performance data and raw log data from an information technology environment |
US10614132B2 (en) | 2013-04-30 | 2020-04-07 | Splunk Inc. | GUI-triggered processing of performance data and log data from an information technology environment |
US10698937B2 (en) | 2017-12-13 | 2020-06-30 | Microsoft Technology Licensing, Llc | Split mapping for dynamic rendering and maintaining consistency of data processed by applications |
CN112053758A (en) * | 2020-08-27 | 2020-12-08 | 北京颢云信息科技股份有限公司 | Intelligent construction and optimization method for single disease database |
US10997191B2 (en) | 2013-04-30 | 2021-05-04 | Splunk Inc. | Query-triggered processing of performance data and log data from an information technology environment |
US11550751B2 (en) | 2016-11-18 | 2023-01-10 | Microsoft Technology Licensing, Llc | Sequence expander for data entry/information retrieval |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5404506A (en) * | 1985-03-27 | 1995-04-04 | Hitachi, Ltd. | Knowledge based information retrieval system |
EP0657828A1 (en) * | 1993-12-06 | 1995-06-14 | Matsushita Electric Industrial Co., Ltd. | An apparatus and a method for retrieving image objects |
-
1997
- 1997-04-04 WO PCT/IB1997/000748 patent/WO1997038376A2/en not_active Application Discontinuation
- 1997-04-04 JP JP9532080A patent/JP2000507008A/en active Pending
- 1997-04-04 EP EP97925221A patent/EP0934569A2/en not_active Withdrawn
- 1997-04-04 CA CA002250694A patent/CA2250694A1/en not_active Abandoned
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5404506A (en) * | 1985-03-27 | 1995-04-04 | Hitachi, Ltd. | Knowledge based information retrieval system |
EP0657828A1 (en) * | 1993-12-06 | 1995-06-14 | Matsushita Electric Industrial Co., Ltd. | An apparatus and a method for retrieving image objects |
Cited By (78)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP1104568A4 (en) * | 1998-07-15 | 2006-04-12 | Amazon Com Inc | System and method for correcting spelling errors in search queries |
US7996398B2 (en) | 1998-07-15 | 2011-08-09 | A9.Com, Inc. | Identifying related search terms based on search behaviors of users |
US7840577B2 (en) | 1998-07-15 | 2010-11-23 | A9.Com, Inc. | Search query processing to identify related search terms and to correct misspellings of search terms |
EP1104568A1 (en) * | 1998-07-15 | 2001-06-06 | Amazon.Com. | System and method for correcting spelling errors in search queries |
US7444324B2 (en) | 1998-07-15 | 2008-10-28 | A9.Com, Inc. | Search query processing to identify search string corrections that reflect past search query submissions of users |
WO2001067280A2 (en) * | 2000-03-08 | 2001-09-13 | Accenture Llp | Method for a knowledge model |
WO2001067280A3 (en) * | 2000-03-08 | 2003-12-24 | Accenture Llp | Method for a knowledge model |
US7350138B1 (en) * | 2000-03-08 | 2008-03-25 | Accenture Llp | System, method and article of manufacture for a knowledge management tool proposal wizard |
US6957205B1 (en) | 2000-03-08 | 2005-10-18 | Accenture Llp | Knowledge model-based indexing of information |
US7099854B2 (en) | 2000-03-08 | 2006-08-29 | Accenture Llp | Knowledgebase framework system |
US6842730B1 (en) | 2000-06-22 | 2005-01-11 | Hapax Limited | Method and system for information extraction |
US7657425B2 (en) | 2000-06-22 | 2010-02-02 | Hapax Limited | Method and system for information extraction |
WO2001098946A1 (en) * | 2000-06-22 | 2001-12-27 | Hapax Information Systems Ab | Method and system for information extraction |
US7194406B2 (en) | 2000-06-22 | 2007-03-20 | Hapax Limited | Method and system for information extraction |
US7072847B2 (en) * | 2000-08-25 | 2006-07-04 | Jonas Ulenas | Method and apparatus for obtaining consumer product preferences through product selection and evaluation |
US6735560B1 (en) * | 2001-01-31 | 2004-05-11 | International Business Machines Corporation | Method of identifying members of classes in a natural language understanding system |
US7979314B2 (en) | 2001-08-23 | 2011-07-12 | Jonas Ulenas | Method and apparatus for obtaining consumer product preferences through interactive product selection and evaluation |
US6996774B2 (en) | 2002-02-12 | 2006-02-07 | Accenture Global Services Gmbh | Display of data element indicia based on data types |
US7089188B2 (en) * | 2002-03-27 | 2006-08-08 | Hewlett-Packard Development Company, L.P. | Method to expand inputs for word or document searching |
GB2417115A (en) * | 2002-09-27 | 2006-02-15 | Hewlett Packard Development Co | Managing synonymic searching and ranking results |
GB2393541A (en) * | 2002-09-27 | 2004-03-31 | Hewlett Packard Development Co | Method for management of synonymic searching |
EP1411448A3 (en) * | 2002-10-17 | 2007-12-05 | Matsushita Electric Industrial Co., Ltd. | Data searching apparatus |
EP1411448A2 (en) * | 2002-10-17 | 2004-04-21 | Matsushita Electric Industrial Co., Ltd. | Data searching apparatus |
US7853556B2 (en) | 2003-09-12 | 2010-12-14 | Accenture Global Services Limited | Navigating a software project respository |
US8356036B2 (en) | 2005-02-04 | 2013-01-15 | Accenture Global Services | Knowledge discovery tool extraction and integration |
US7904411B2 (en) | 2005-02-04 | 2011-03-08 | Accenture Global Services Limited | Knowledge discovery tool relationship generation |
US8660977B2 (en) | 2005-02-04 | 2014-02-25 | Accenture Global Services Limited | Knowledge discovery tool relationship generation |
US8010581B2 (en) | 2005-02-04 | 2011-08-30 | Accenture Global Services Limited | Knowledge discovery tool navigation |
US7716201B2 (en) | 2006-08-10 | 2010-05-11 | Yahoo! Inc. | Method and apparatus for reconstructing a search query |
US8209317B2 (en) | 2006-08-10 | 2012-06-26 | Yahoo! Inc. | Method and apparatus for reconstructing a search query |
US8046347B2 (en) | 2006-08-10 | 2011-10-25 | Yahoo! Inc. | Method and apparatus for reconstructing a search query |
US9747316B2 (en) | 2006-10-05 | 2017-08-29 | Splunk Inc. | Search based on a relationship between log data and data from a real-time monitoring environment |
US11561952B2 (en) | 2006-10-05 | 2023-01-24 | Splunk Inc. | Storing events derived from log data and performing a search on the events and data that is not log data |
US10977233B2 (en) | 2006-10-05 | 2021-04-13 | Splunk Inc. | Aggregating search results from a plurality of searches executed across time series data |
US8990184B2 (en) | 2006-10-05 | 2015-03-24 | Splunk Inc. | Time series search engine |
US9002854B2 (en) | 2006-10-05 | 2015-04-07 | Splunk Inc. | Time series search with interpolated time stamp |
US9514175B2 (en) | 2006-10-05 | 2016-12-06 | Splunk Inc. | Normalization of time stamps for event data |
US9594789B2 (en) | 2006-10-05 | 2017-03-14 | Splunk Inc. | Time series search in primary and secondary memory |
EP2074505A4 (en) * | 2006-10-05 | 2010-01-13 | Splunk Inc | Time series search engine |
US9922067B2 (en) | 2006-10-05 | 2018-03-20 | Splunk Inc. | Storing log data as events and performing a search on the log data and data obtained from a real-time monitoring environment |
US9922066B2 (en) | 2006-10-05 | 2018-03-20 | Splunk Inc. | Aggregation and display of search results from multi-criteria search queries on event data |
US9922065B2 (en) | 2006-10-05 | 2018-03-20 | Splunk Inc. | Determining timestamps to be associated with events in machine data |
US9928262B2 (en) | 2006-10-05 | 2018-03-27 | Splunk Inc. | Log data time stamp extraction and search on log data real-time monitoring environment |
US9996571B2 (en) | 2006-10-05 | 2018-06-12 | Splunk Inc. | Storing and executing a search on log data and data obtained from a real-time monitoring environment |
US11947513B2 (en) | 2006-10-05 | 2024-04-02 | Splunk Inc. | Search phrase processing |
US10216779B2 (en) | 2006-10-05 | 2019-02-26 | Splunk Inc. | Expiration of persistent data structures that satisfy search queries |
US10891281B2 (en) | 2006-10-05 | 2021-01-12 | Splunk Inc. | Storing events derived from log data and performing a search on the events and data that is not log data |
US10242039B2 (en) | 2006-10-05 | 2019-03-26 | Splunk Inc. | Source differentiation of machine data |
US10255312B2 (en) | 2006-10-05 | 2019-04-09 | Splunk Inc. | Time stamp creation for event data |
US10262018B2 (en) | 2006-10-05 | 2019-04-16 | Splunk Inc. | Application of search policies to searches on event data stored in persistent data structures |
EP3493074A1 (en) * | 2006-10-05 | 2019-06-05 | Splunk Inc. | Time series search engine |
US11550772B2 (en) | 2006-10-05 | 2023-01-10 | Splunk Inc. | Time series search phrase processing |
US11537585B2 (en) | 2006-10-05 | 2022-12-27 | Splunk Inc. | Determining time stamps in machine data derived events |
US11526482B2 (en) | 2006-10-05 | 2022-12-13 | Splunk Inc. | Determining timestamps to be associated with events in machine data |
US10747742B2 (en) | 2006-10-05 | 2020-08-18 | Splunk Inc. | Storing log data and performing a search on the log data and data that is not log data |
US11249971B2 (en) | 2006-10-05 | 2022-02-15 | Splunk Inc. | Segmenting machine data using token-based signatures |
US10678767B2 (en) | 2006-10-05 | 2020-06-09 | Splunk Inc. | Search query processing using operational parameters |
US11144526B2 (en) | 2006-10-05 | 2021-10-12 | Splunk Inc. | Applying time-based search phrases across event data |
US10740313B2 (en) | 2006-10-05 | 2020-08-11 | Splunk Inc. | Storing events associated with a time stamp extracted from log data and performing a search on the events and data that is not log data |
US7953687B2 (en) | 2006-11-13 | 2011-05-31 | Accenture Global Services Limited | Knowledge discovery system with user interactive analysis view for analyzing and generating relationships |
US7765176B2 (en) | 2006-11-13 | 2010-07-27 | Accenture Global Services Gmbh | Knowledge discovery system with user interactive analysis view for analyzing and generating relationships |
US10592522B2 (en) | 2013-04-30 | 2020-03-17 | Splunk Inc. | Correlating performance data and log data using diverse data stores |
US10346357B2 (en) | 2013-04-30 | 2019-07-09 | Splunk Inc. | Processing of performance data and structure data from an information technology environment |
US10019496B2 (en) | 2013-04-30 | 2018-07-10 | Splunk Inc. | Processing of performance data and log data from an information technology environment by using diverse data stores |
US11782989B1 (en) | 2013-04-30 | 2023-10-10 | Splunk Inc. | Correlating data based on user-specified search criteria |
US10997191B2 (en) | 2013-04-30 | 2021-05-04 | Splunk Inc. | Query-triggered processing of performance data and log data from an information technology environment |
US10225136B2 (en) | 2013-04-30 | 2019-03-05 | Splunk Inc. | Processing of log data and performance data obtained via an application programming interface (API) |
US11119982B2 (en) | 2013-04-30 | 2021-09-14 | Splunk Inc. | Correlation of performance data and structure data from an information technology environment |
US10877986B2 (en) | 2013-04-30 | 2020-12-29 | Splunk Inc. | Obtaining performance data via an application programming interface (API) for correlation with log data |
US10318541B2 (en) | 2013-04-30 | 2019-06-11 | Splunk Inc. | Correlating log data with performance measurements having a specified relationship to a threshold value |
US10614132B2 (en) | 2013-04-30 | 2020-04-07 | Splunk Inc. | GUI-triggered processing of performance data and log data from an information technology environment |
US10353957B2 (en) | 2013-04-30 | 2019-07-16 | Splunk Inc. | Processing of performance data and raw log data from an information technology environment |
US11550751B2 (en) | 2016-11-18 | 2023-01-10 | Microsoft Technology Licensing, Llc | Sequence expander for data entry/information retrieval |
US10698937B2 (en) | 2017-12-13 | 2020-06-30 | Microsoft Technology Licensing, Llc | Split mapping for dynamic rendering and maintaining consistency of data processed by applications |
US11126648B2 (en) | 2017-12-13 | 2021-09-21 | Microsoft Technology Licensing, Llc | Automatically launched software add-ins for proactively analyzing content of documents and soliciting user input |
US11061956B2 (en) | 2017-12-13 | 2021-07-13 | Microsoft Technology Licensing, Llc | Enhanced processing and communication of file content for analysis |
US10929455B2 (en) | 2017-12-13 | 2021-02-23 | Microsoft Technology Licensing, Llc | Generating an acronym index by mining a collection of document artifacts |
CN112053758A (en) * | 2020-08-27 | 2020-12-08 | 北京颢云信息科技股份有限公司 | Intelligent construction and optimization method for single disease database |
Also Published As
Publication number | Publication date |
---|---|
WO1997038376A3 (en) | 1997-12-04 |
EP0934569A2 (en) | 1999-08-11 |
CA2250694A1 (en) | 1997-10-16 |
JP2000507008A (en) | 2000-06-06 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO1997038376A2 (en) | A system, software and method for locating information in a collection of text-based information sources | |
US6161084A (en) | Information retrieval utilizing semantic representation of text by identifying hypernyms and indexing multiple tokenized semantic structures to a same passage of text | |
US5331556A (en) | Method for natural language data processing using morphological and part-of-speech information | |
US7567902B2 (en) | Generating speech recognition grammars from a large corpus of data | |
US8041697B2 (en) | Semi-automatic example-based induction of semantic translation rules to support natural language search | |
JP5243167B2 (en) | Information retrieval system | |
US7734623B2 (en) | Semantics-based method and apparatus for document analysis | |
US20050080613A1 (en) | System and method for processing text utilizing a suite of disambiguation techniques | |
JP2011118689A (en) | Retrieval method and system | |
Pandey et al. | An unsupervised Hindi stemmer with heuristic improvements | |
Nakanishi et al. | Probabilistic models for disambiguation of an HPSG-based chart generator | |
US20220229998A1 (en) | Lookup source framework for a natural language understanding (nlu) framework | |
US20220229990A1 (en) | System and method for lookup source segmentation scoring in a natural language understanding (nlu) framework | |
US20220229986A1 (en) | System and method for compiling and using taxonomy lookup sources in a natural language understanding (nlu) framework | |
US20220237383A1 (en) | Concept system for a natural language understanding (nlu) framework | |
Ahkouk et al. | Human language question to sql query using deep learning | |
Rishel et al. | Augmentation of a term/document matrix with part-of-speech tags to improve accuracy of latent semantic analysis. | |
JP2005025555A (en) | Thesaurus construction system, thesaurus construction method, program for executing the method, and storage medium with the program stored thereon | |
JP4635585B2 (en) | Question answering system, question answering method, and question answering program | |
JP3176750B2 (en) | Natural language translator | |
JP3419748B2 (en) | Dictionary creation device and method, and recording medium recording dictionary creation program | |
Arun | Statistical parsing of the French treebank | |
Asubiaro | An Analysis of the Structure of Index Terms for Yoruba Texts | |
JPH11120193A (en) | Method and device for retrieving natural sentence and storage medium recording natural sentence retrieval program | |
JPH0320866A (en) | Text base retrieval system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AK | Designated states |
Kind code of ref document: A2 Designated state(s): CA JP |
|
AL | Designated countries for regional patents |
Kind code of ref document: A2 Designated state(s): AT BE CH DE DK ES FI FR GB GR IE IT LU MC NL PT SE |
|
AK | Designated states |
Kind code of ref document: A3 Designated state(s): CA JP |
|
AL | Designated countries for regional patents |
Kind code of ref document: A3 Designated state(s): AT BE CH DE DK ES FI FR GB GR IE IT LU MC NL PT SE |
|
DFPE | Request for preliminary examination filed prior to expiration of 19th month from priority date (pct application filed before 20040101) | ||
121 | Ep: the epo has been informed by wipo that ep was designated in this application | ||
ENP | Entry into the national phase in: |
Ref country code: CA Ref document number: 2250694 Kind code of ref document: A Format of ref document f/p: F Ref document number: 2250694 Country of ref document: CA |
|
WWE | Wipo information: entry into national phase |
Ref document number: 1997925221 Country of ref document: EP |
|
WWP | Wipo information: published in national office |
Ref document number: 1997925221 Country of ref document: EP |
|
WWW | Wipo information: withdrawn in national office |
Ref document number: 1997925221 Country of ref document: EP |