EP0597630A1
Expired / Lapsed
EP
⏱️ 0.0 years remaining
Method for resolution of natural-language queries against full-text databases
Assignee: Conquest Software Inc
•
Filing: 1993-11-04
•
Granted: —
•
Expires: 2013-11-04
Abstract
Abstract The method of the present invention combines concept searching, document ranking, high speed and efficiency, browsing capabilities, "intelligent" hypertext, document routing, and summarization (machine abstracting) in an easy-to-use implementation. The method of the present invention also offers Boolean and statistical query options. The method of the present invention is based upon "concept indexing" (an index of "word senses" rather than just words.) It builds its concept index from a "semantic network" of word relationships with word definitions drawn from one or more standard human-language dictionaries. During query, users may select the meaning of a word from the dictionary during query construction, or may allow the method to disambiguate words based on semantic and statistical evidence of meaning. This results in a measurable improvement in precision and recall. Results of searching are retrieved and displayed in ranked order. The ranking process is more sophisticated than prior art systems providing ranking because it takes linguistics and concepts, as well as statistics into account.
Inventors
- 👤 Edwin R. Conquest Software Inc. Addison
- 👤 Arden S. Conquest Software Inc. Blair
- 👤 Paul E. Conquest Software Inc. Nelson
- 👤 Thomas Conquest Software Inc. Schwartz
Bibliographic Details
Application Number
—
Jurisdiction
EP
Data Source
google_patents
Last Synchronized
2026-10-25 09:22
CPC Classifications
G; G06; G06F; G06F16/00; G06F16/30; G06F16/36; G06F16/374; G06F16/33; G06F16/3331; G06F16/3332; G06F16/3334; G06F16/334; G06F16/3344; G06F16/34; G06F16/345; G06F40/00; G06F40/20; G06F40/253
Claims Analysis
Total Claims: 18
Claim 1
Independent Claim
117 words
A method for searching a computer database containing one or more documents comprised of symbols representative of information expressed in a language which is understandable to human users comprising the steps of: (a) accepting a query comprising one or more symbols representative of information comprising one or more terms; (b) determining one or more likely meanings for each term in said query; (c) identifying in rank order one or more of said likely meanings in said database; (d) identifying in rank order documents within said database which contain at least one said likely meanings identified in step (c); (e) storing an indication of the identities of said identified individual documents in the memory of a digital computer.
Claim 2
Dependent Claim
25 words
The method of Claim 1 wherein said step (a) of accepting a query comprises entry of a sentence in English or another natural human language.
Claim 3
Dependent Claim
54 words
The method of Claim 1 wherein said step (a) of accepting a query comprises accepting a document comprised of symbols representative of information expressed in a language which is understandable to human users, said document containing information which intended by a user to be similar to the information contained in said identified individual documents.
Claim 4
Dependent Claim
49 words
The method of Claim 1 wherein said step (a) of accepting a query comprises identification of one or more subsets of symbols representative of information expressed in a language which is understandable to human users, which comprise one or more subsets of one or more documents within said database.
Claim 5
Dependent Claim
56 words
The method of Claim 1 wherein each identified term is compared to a predetermined list of terms which occur in linguistic idioms, and if a term is determined to exist in said list, then storing that term, and an identification of the idiom or idioms in which it occurs in the memory of a digital computer.
Claim 6
Dependent Claim
81 words
The method of Claim 5 wherein said step (b) of determining one or more likely meanings for each term in said query further comprises looking up each identifed term in a machine-readable dictionary, and retrieving from said dictionary information comprising at least one of the following: meanings of the term, the parts of speech of each such meaning, pointer data structures into an associated semantic network for each such meaning, and information about the use of the term in linguistic idioms.
Claim 7
Dependent Claim
27 words
The method of Claim 5 wherein the syntax of a multi-term query is analyzed to reduce the number of likely meanings of each term in said query.
Claim 8
Dependent Claim
38 words
The method of Claim 5 wherein the likely meanings of each term in said query are evaluated by spreading activation through an associative network of term meanings to accumulate statistical evidence of the likelihood of particular terms meanings.
Claim 9
Dependent Claim
33 words
The method of Claim 8 wherein said associative network of term meanings is derived by processing of machine-readable sources of linguistic information, using said linguistic information to develop association links and link strengths.
Claim 10
Dependent Claim
61 words
The method of Claim 1 wherein step (c) of identifying in rank order one or more of said likely meanings in said database comprises a ranking according to at least one of the following criteria: inverse document frequency, syntactic position, part of speech, application of a predetermined concept tree, part of speech, predetermined ranking, and explicit selection by a human user.
Claim 11
Dependent Claim
62 words
The method of Claim 1 wherein step (d) of identifying documents further comprises the steps of: (d1) identifying in rank order sets of documents within said database containing one or more documents which contain at least one said likely meanings identified in step (c); and (d2) successively identifying smaller sets containing documents until a predetermined number of individual documents have been identified.
Claim 12
Independent Claim
136 words
A method for searching a computer database containing one or more documents comprised of symbols representative of information expressed in a language which is understandable to human users comprising the steps of: (a) accepting a query comprising one or more symbols representative of information comprising one or more terms; (b) determining one or more likely meanings for each term in said query; (c) identifying in rank order one or more of said likely meanings in said database; (d) identifying additional likely meanings which are related to said likely meanings identified in step (c); (e) identifying in rank order one or more documents which contain at least one said likely meanings identified in step (c) or (d); and (f) storing an indication of the identities of said identified individual documents in the memory of a digital computer.
Claim 13
Independent Claim
141 words
A method for hypertext searching a computer database containing one or more documents comprised of sections, said sections comprised of symbols representative of information expressed in a language which is understandable to human users comprising the steps of: (a) accepting a query comprising one or more symbols representative of information comprising one or more terms, said query comprising an identification of one or more sections of a selected document within said database; (b) determining one or more likely meanings for each term in said query; (c) identifying in rank order one or more of said likely meanings in said database; (d) identifying in rank order one or more sections which contain at least one said likely meanings identified in step (c); and (f) storing an indication of the identities of said identified individual sections in the memory of a digital computer.
Claim 14
Independent Claim
131 words
A method for constructing a combination associative network of term meanings and machine-readable dictionary from a plurality of machine-readable linguistic databases comprising the following steps: (a) identifying root words and their associated meanings; (b) identifying non-root words and their associated meanings, and identifying their relationships to the root words identified in step (a); (c) identifying descriptive phrases and idioms and their associated meanings, and identifying their relationships to the root words identified in step (a); (d) identifying link relationships among the root words, non-root words, descriptive phrases, idioms, and their associated meanings identified in steps (a-c); (e) determining link strengths for each link relationship identified in step (d); and (f) storing said root words, non-root words, descriptive phrases, idioms, meanings, links, and link strengths in the memory of a digital computer.
Claim 15
Independent Claim
83 words
A method for real-time characterization of source documents comprised of a plurality of symbols representative of information expressed in a language which is understandable to human users comprising the steps of: (a) identifying terms within said source document; (b) searching a database containing documents comprised of one or more terms and enhancement information associated therewith; (c) identifying documents within said database containing one or more of said identifed terms; (d) associating the enhancement information associated with said identfied documents with said source document.
Claim 16
Independent Claim
112 words
A method for enhancing the content of a document comprised of symbols representative of information expressed in a language which is understandable to human users comprising the steps of: (a) identifying one or more terms comprised of one or more symbols within said document; (b) determining one or more likely meanings for each term in said document; (c) identifying in rank order one or more of said likely meanings of said identified terms; (d) optionally identifying additional likely meanings which are related to said likely meanings identified in step (c); and (e) storing said document, said identified likely meanings, and said identified additional likely meanings in the memory of a digital computer.
Claim 17
Dependent Claim
132 words
The method of Claim 16 wherein said database contains documents comprised of symbols representative of information expressed in a language which is understandable to human users, and wherein said database is constructed according to a method comprising the steps of: (a) identifying one or more terms comprised of one or more symbols within each of said documents; (b) determining one or more likely meanings for each term in each of said documents; (c) identifying in rank order one or more of said likely meanings of said identified terms; (d) optionally identifying additional likely meanings which are related to said likely meanings identified in step (c); and (e) storing each of said documents, said identified likely meanings, and said identified additional likely meanings in the memory of a digital computer as said database.
Claim 18
Independent Claim
114 words
A method for indexing a document comprised of symbols representative of information expressed in a language which is understandable to human users comprising the steps of: (a) identifying one or more terms comprised of one or more symbols within said document; (b) determining one or more likely meanings for each term in said document; (c) identifying in rank order one or more of said likely meanings of said identified terms; (d) determining the informational value of each of said likely meanings, and discarding those likely meanings having an informational value which is less than a predetermined value; and (e) storing said document and said identified likely meanings in the memory of a digital computer.
References Cited (Backward Citations)
USPTO Annuity & Maintenance Windows
Maintenance schedule is calculated for granted US utility patents.