# US 2015/0310005 A1 · Method, system and software for searching, identifying, retrieving and presenting electronic documents

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2015/0310005 A1
- Document ID: US20150310005A1
- Status: Published application
- Kind: A1
- Application number: 14/673,801
- Filed: 2015-03-30
- Published: 2015-10-29
- Earliest claimed: 2014-03-29
- Assignees: Thomson Reuters Global Resources
- Inventors: Raphael Shmuel Ryger; Ekaterina Suvorova
- Web page: https://patentgrep.com/patent/US20150310005A1

## CPC

- G06F 16/93: Document management systems
- G06F 16/24578: {using ranking}
- G06F 16/3334: {Selection or weighting of terms from queries, including natural language queries}
- G06F 16/9535: Search customisation based on user profiles and personalisation
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

The present invention provides a method and system for identifying, retrieving and presenting electronic documents responsive to user queries. The three distinct inventive concepts are relevancy ranking of responsive documents based on component query technique; cross-lingual searching; and search expansion using analytics of initial results to derive and generate a modified query. Each of these inventions enhances document search and retrieval systems and the three solutions may be used separately or in any combination. The three inventions apply in layers above an underlying search system, controlling the submission of requests to the underlying system in support of received search requests, typically originating with an end user. Invention (I) involves use of either of two general types of enhanced OR operator that may be offered by the underlying search system. Processing of the content domain informs automated generation of elaborated queries using these operators for submitting to such an underlying system so as to yield improved relevance ranking. Invention (II) introduces a cross-lingual search-term suggestion capability driven by the structured data being searched, requiring no dictionaries or thesauri and no rule-based or statistical language modeling. Invention (III) introduces a semantic-expansion capability specifically utilizing the availability of language independent fields in the data being searched with natural language query terms. This invention is enhanced by incorporating the preceding two inventions.

## Claims

### Claim 1

A computer-implemented method of improving performance of a generic text-search engine by use of a search-controller to provide cross-lingual searching in a computer-implemented search of a corpus of documents, the method comprising:
a. operating at the search-controller level to mediate between a request function and an underlying generic text-search system and electronically receiving a query containing a set of user-supplied search terms in a source language from the request function;
b. submitting the set of user-supplied search terms in the source language to the underlying search engine and returning a set of responsive documents by matching the set of user-supplied search terms with terms in a set of source language fields associated with the set of responsive documents and having a set of values that are relatively short and keyword dense in comparison to the associated document;
c. statistically analyzing the set of values in a set of target language fields associated with the set of responsive documents, the target language fields being relatively short and keyword dense in comparison to the associated document; and
d. generating a set of suggested search terms in a target language other than the source language.

### Claim 2 (depends on 1)

The method of claim 1 further comprising:
automatically choosing by the search controller the set of source-language fields having values that are relatively short and keyword dense in comparison to the associated document; and
automatically choosing by the search controller the set of target-language fields having values that are relatively short and keyword dense in comparison to the associated document.

### Claim 3 (depends on 1)

The method of claim 1 further comprising scoring the set of suggested search terms based at least in part on a set of usability criteria, with each suggested search term scored for its usability as a search term against one of the set of target language fields.

### Claim 4 (depends on 1)

The method of claim 1 further comprising selecting a set of suggested terms based on a set of threshold criteria for use in a search against one or more of the set of target-language fields.

### Claim 5 (depends on 1)

The method of claim 1 wherein the selected set of suggested terms is used in searching the target language fields in addition to or instead of using the set of user-supplied search terms to search the set of source language fields to find documents of interest.

### Claim 6 (depends on 1)

The method of claim 1 wherein the short, keyword dense source language fields and target language fields comprise a combination of at least two from the set of title, abstract, and keywords fields.

### Claim 7 (depends on 1)

The method of claim 1 wherein statistically analyzing the set of values in a set of target language fields includes considering the frequency of values in the target-language fields of the documents returned by the generic text search engine from the search for the set of user-supplied search terms in the set of source language fields, and further considers the distribution of field values in the set of target language fields or of the corpus of documents.

### Claim 8 (depends on 7)

The method of claim 7 wherein statistically analyzing the set of values in a set of target language fields includes downwardly adjusting scores of terms where such terms are frequent in both the set of target language fields of the set of responsive documents and in the set of target language fields of the documents of the corpus in general.

### Claim 9 (depends on 1)

The method of claim 1 wherein one or more configuration parameters are set to reduce the processing involved in the search and the quantity of analytical data generated and communicated to users.

### Claim 10 (depends on 9)

The method of claim 9 wherein the configuration parameters are one or more of numerical thresholds and specification of subportions of the corpus to be accessed.

### Claim 11 (depends on 1)

The method of claim 1 further comprising providing a user interface for presenting and receiving responses to suggested search terms whereby a user may select or de-select from a set of presented suggested search terms for exectution by the generic search engine.

### Claim 12 (depends on 11)

The method of claim 11 wherein the user may request additional suggested search terms by iteratively invoking the term-suggestion process and selecting/de-selecting from a set of returned items as the “seed” for solicitation of further search-term suggestions.

### Claim 13 (depends on 12)

The method of claim 12 wherein the iterative rounds of suggestions each being presented in its own panel of scored suggested search-term items, resulting in a cascade of suggestion panels.

### Claim 14 (depends on 1)

The method of claim 1 further comprising providing a user interface for enabling a user to indicate a particular search term in a text-search query being prepared as in need of search term suggestion.

### Claim 15 (depends on 14)

The method of claim 14 further comprising upon receiving an input from a user requesting search term suggestions, automatically generating for display a panel comprising a set of suggested search terms.

### Claim 16 (depends on 15)

The method of claim 15 wherein each suggested search term presented in the panel being selectable for use within the text-search query being prepared by the user.

### Claim 17 (depends on 1)

The method of claim 1 further comprising:
d. given the received query comprising the user-supplied set of search terms, determining a topic based on the set of search terms and identifying documents related to the topic irrespective of whether the identified documents contain any of the set of search terms, the step of identifying documents comprising:
e. performing a narrow search-term expansion by determining alternative terms to be combined with the submitted set of search terms to create a narrowly expanded search term set for use by the search engine, narrow search-term expansion adapted to result in an initial broadening of the sampling-search step (d);
f. performing a sampling-search using the narrowly expanded search term set determined in step (e); and
g. retrieving data from a set of sample fields as a sample set of the identified complete set of responsive documents, the set of sample fields being preselected to be language independent but strongly topic correlated.

### Claim 18 (depends on 17)

The method of claim 17 further comprising:
h. statistical analysis of the distribution of values in the retrieved content-correlated, language-independent fields;
i. computation of a “signature” of the responsive documents in terms of values of the respective fields per the analysis in step (h), the signature incorporating quantifiers monotonically related to the observed prevalence of the values in their respective fields as revealed in that analysis;
j. submission, to the underlying search system, of a query that is a weighted-OR of criteria stipulating occurrence of the top values observed in the analysis in step (h), the weights on the values being the computed quantifiers of the signature from step (i), together with criteria stipulating occurrence of the terms from the narrowly expanded arrays of search terms determined in step (e); and
k. collection of identifiers of the responsive documents from step (j), each with its relevance score as computed by the underlying search system, in order of decreasing relevance score, cutting off at a point determined by configuration parameters.

### Claim 19 (depends on 17)

The method of claim 17 wherein the set of sample fields includes one or more from the set consisting of classification-code fields, cited-document fields, standardized document identifiers, and standardized keyword fields.

### Claim 20

A computer-implemented method of improving performance of a generic text-search engine by use of a search-controller to provide cross-lingual searching in a computer-implemented search of a corpus of documents, the method comprising:
a. operating at the search-controller level to mediate between a request function and an underlying generic text-search system and electronically receiving a query containing a set of user-supplied search terms in a source language from the request function;
b. submitting the set of user-supplied search terms in the source language to the underlying search engine and returning a set of responsive documents by matching the set of user-supplied search terms with terms in a set of source language fields associated with the set of responsive documents and having a set of values that are relatively short and keyword dense in comparison to the associated document;
c. statistically analyzing the set of values in a set of target language fields associated with the set of responsive documents, the target language fields being relatively short and keyword dense in comparison to the associated document; and
d. generating a set of suggested search terms in a target language that is the same as the source language.

### Claim 21

A facility including a computer-based search system configured to improve operational search performance of the search system to provide cross-lingual searching in a computer-implemented search of a corpus of documents, the facility comprising:
a generic text-search engine;
a search-controller in communication with the generic text-search engine and an applications services layer connected to a communications network for receiving a search query, the search-controller configured to mediate between a request function and the generic text-search engine and to electronically receive a query containing a set of user-supplied search terms in a source language from the request function, the search-controller comprising:
a search federator adapted to submit the set of user-supplied search terms in the source language to the generic text-search engine and return a set of responsive documents by matching the set of user-supplied search terms with terms in a set of source language fields associated with the set of responsive documents and having a set of values that are relatively short and keyword dense in comparison to the associated document;
a search-results analyzer adapted to statistically analyze the set of values in a set of target language fields associated with the set of responsive documents, the target language fields being relatively short and keyword dense in comparison to the associated document; and
a cross-lingual search-term suggester adapted to generate a set of suggested search terms in a target language other than the source language.

### Claim 22 (depends on 21)

The facility of claim 21 further comprising:
automatically choosing by the search controller the set of source-language fields having values that are relatively short and keyword dense in comparison to the associated document; and
automatically choosing by the search controller the set of target-language fields having values that are relatively short and keyword dense in comparison to the associated document.

### Claim 23 (depends on 21)

The facility of claim 21 further comprising scoring the set of suggested search terms based at least in part on a set of usability criteria, with each suggested search term scored for its usability as a search term against one of the set of target language fields.

### Claim 24 (depends on 21)

The facility of claim 21 further comprising selecting a set of suggested terms based on a set of threshold criteria for use in a search against one or more of the set of target-language fields.

### Claim 25 (depends on 21)

The facility of claim 21 wherein the selected set of suggested terms is used in searching the target language fields in addition to or instead of using the set of user-supplied search terms to search the set of source language fields to find documents of interest.

### Claim 26 (depends on 21)

The facility of claim 21 wherein the short, keyword dense source language fields and target language fields comprise a combination of at least two from the set of title, abstract, and keywords fields.

### Claim 27 (depends on 21)

The facility of claim 21 wherein statistically analyzing the set of values in a set of target language fields includes considering the frequency of values in the target-language fields of the documents returned by the generic text search engine from the search for the set of user-supplied search terms in the set of source language fields, and further considers the distribution of field values in the set of target language fields or of the corpus of documents.

### Claim 28 (depends on 21)

The facility of claim 21 wherein statistically analyzing the set of values in a set of target language fields includes downwardly adjusting scores of terms where such terms are frequent in both the set of target language fields of the set of responsive documents and in the set of target language fields of the documents of the corpus in general.

### Claim 29 (depends on 21)

The facility of claim 21 wherein one or more configuration parameters are set to reduce the processing involved in the search and the quantity of analytical data generated and communicated to users.

### Claim 30 (depends on 21)

The facility of claim 21 wherein the configuration parameters are one or more of numerical thresholds and specification of subportions of the corpus to be accessed.

### Claim 31 (depends on 21)

The facility of claim 21 further comprising providing a user interface for presenting and receiving responses to suggested search terms whereby a user may select or de-select from a set of presented suggested search terms for exectution by the generic search engine.

### Claim 32 (depends on 21)

The facility of claim 21 wherein the user may request additional suggested search terms by iteratively invoking the term-suggestion process and selecting/de-selecting from a set of returned items as the “seed” for solicitation of further search-term suggestions.

### Claim 33 (depends on 21)

The facility of claim 21 wherein the iterative rounds of suggestions each being presented in its own panel of scored suggested search-term items, resulting in a cascade of suggestion panels.

### Claim 34 (depends on 21)

The facility of claim 21 further comprising providing a user interface for enabling a user to indicate a particular search term in a text-search query being prepared as in need of search term suggestion.

### Claim 35 (depends on 34)

The facility of claim 34 further comprising upon receiving an input from a user requesting search term suggestions, automatically generating for display a panel comprising a set of suggested search terms.

### Claim 36 (depends on 35)

The facility of claim 35 wherein each suggested search term presented in the panel being selectable for use within the text-search query being prepared by the user.

### Claim 37 (depends on 21)

The facility of claim 21 further comprising:
d. given the received query comprising the user-supplied set of search terms, determining a topic based on the set of search terms and identifying documents related to the topic irrespective of whether the identified documents contain any of the set of search terms, the step of identifying documents comprising:
e. performing a narrow search-term expansion by determining alternative terms to be combined with the submitted set of search terms to create a narrowly expanded search term set for use by the search engine, narrow search-term expansion adapted to result in an initial broadening of the sampling-search step (d);
f. performing a sampling-search using the narrowly expanded search term set determined in step (e); and
g. retrieving data from a set of sample fields as a sample set of the identified complete set of responsive documents, the set of sample fields being preselected to be language independent but strongly topic correlated.

### Claim 38 (depends on 37)

The facility of claim 37 further comprising:
h. statistical analysis of the distribution of values in the retrieved content-correlated, language-independent fields;
i. computation of a “signature” of the responsive documents in terms of values of the respective fields per the analysis in step (h), the signature incorporating quantifiers monotonically related to the observed prevalence of the values in their respective fields as revealed in that analysis;
j. submission, to the underlying search system, of a query that is a weighted-OR of criteria stipulating occurrence of the top values observed in the analysis in step (h), the weights on the values being the computed quantifiers of the signature from step (i), together with criteria stipulating occurrence of the terms from the narrowly expanded arrays of search terms determined in step (e); and
k. collection of identifiers of the responsive documents from step (j), each with its relevance score as computed by the underlying search system, in order of decreasing relevance score, cutting off at a point determined by configuration parameters.

### Claim 39 (depends on 37)

The facility of claim 37 wherein the set of sample fields includes one or more from the set consisting of classification-code fields, cited-document fields, standardized document identifiers, and standardized keyword fields.

### Claim 40

A computer implemented method for recommending a set of search terms to be used in a cross-lingual search, the method comprising:
a. receiving a search term representing a set of characters in a first language;
b. searching a database of documents for documents containing the received search term;
c. in response to the step of searching, identifying a subset of at least one document from the set of documents contained in the database, each document in the subset of at least one document comprising the received search term and a second language set of information;
d. generating a term frequency table based upon a collection of the second language set of information from each document in the subset of at least one document; and
e. based upon the term frequency table, generating a signal associated with a set of recommended search terms to be used in the cross-lingual search;
f. storing the signal; and
g. transmitting the signal.

### Claim 41 (depends on 1)

The method of claim 1 further comprising:
h. in response to the step of transmitting the signal, receiving a second signal indicative of a selected set of search terms from the set of recommended search terms;
i. searching the database for the selected set of search terms;
j. in response to the step of searching the database for the selected set of search terms, identifying a second set of documents and generating a second signal associated with the second set of documents;
k. storing the second signal; and
l. transmitting the second signal.

### Claim 42 (depends on 40)

The method of claim 40 wherein the second language set of information is in English.

### Claim 43 (depends on 42)

The method of claim 42 wherein the first language is a language other than English.
