Extending keyword searching to syntactically and semantically annotated data

• US 20070156669A1
• Filed: 11/16/2006
• Published: 07/05/2007
• Est. Priority Date: 11/16/2005
• Status: Active Grant

• Current Assignee: Fiver LLC
• Original Assignee: VCVC III LLC (Vulcan, Inc.)
• Inventors: Dhillon, Navdeep, Liang, Jisheng, Nguyen, Thien, Koperski, Krzysztof, Tusk, Carsten, Brown, Matthew, Marchisio, Giovanni
• Granted Patent: US 8,856,096 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/3344 using natural language analysis
• CPC Class Codes: G06F 16/36 Creation of semantic tools, e.g. ontology or thesauri
• CPC Class Codes: G06F 16/93 Document management systems
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques

Abstract

Methods and systems for extending keyword searching techniques to syntactically and semantically annotated data are provided. Example embodiments provide a Syntactic Query Engine (“SQE”) that parses, indexes, and stores a data set as an enhanced document index with document terms as well as information pertaining to the grammatical roles of the terms and ontological and other semantic information. In one embodiment, the enhanced document index is a form of term-clause index, that indexes terms and syntactic and semantic annotations at the clause level. The enhanced document index permits the use of a traditional keyword search engine to process relationship queries as well as to process standard document level keyword searches. This abstract is provided to comply with rules requiring an abstract, and it is submitted with the intention that it will not be used to interpret or limit the scope or meaning of the claims.

Claims

1. A method in a computing system for performing enhanced electronic relationship searches on a corpus of documents, comprising:

receiving an indication of a designated query;

determining one or more matching rules that correspond to the designated query, each rule having associated relationship search specification that indicates a corresponding relationship search;

for each determined rule, determining and executing the corresponding relationship search against the corpus;

for one or more corresponding relationship searches that succeed, forwarding an indication of the corresponding relationship search as a suggested search tip.

2. The method of claim 1 wherein the receiving the indication of the designated query further comprises receiving an indication of one or more components of a relationship query.

3. The method of claim 1 wherein the receiving the indication of the designated query further comprises receiving an indication of more or more keywords.

4. The method of claim 1 wherein the receiving the indication of the designated query comprises receiving an indication of an entity or an entity type.

5. The method of claim 4 wherein the entity or the entity type is a source for an action or for an action type.

6. The method of claim 4 wherein the entity or the entity type is a target for an action or for an action type.

7. The method of claim 1 wherein the receiving the indication of the designated query comprises receiving an indication of an action or an action type.

8. The method of claim 1, further comprising:

for the one or more corresponding relationship searches that succeed,

evaluating results of the corresponding relationship search that succeeds to determine a plurality of specific relationship queries to serve as additional suggested searches; and

forwarding indications of the plurality of specific relationship queries as additional suggested search tips.

9. The method of claim 1 wherein, for the one or more corresponding relationship searches that succeed, the forwarding the indication of the corresponding relationship search as a suggested search tip further comprises:

evaluating results of the corresponding relationship search to determine whether the results meet a set of criteria for presenting as a suggested search tip; and

when it is determined that results meet the criteria, forwarding an indication of the corresponding relationship search as a suggested search tip.

10. The method of claim 1 wherein a corresponding relationship search that succeeds is determined by whether search results are produced from the executing the corresponding relationship search against the corpus.

11. The method of claim 1 wherein the suggested search tip is presented as a navigation tip in a visual user interface.

12. The method of claim 1, further comprising:

caching search results of one or more corresponding relationship searches that succeed; and

when a suggested searched tip is indicated for execution, returning the cached results of the relationship search that corresponds to the suggested search.

13. The method of claim 1 wherein the rules are specified by a configuration file, and further comprising:

parsing the configuration file to determine an associated relationship search specification that indicates a corresponding relationship query for each rule;

for each determined relationship search specification, storing the corresponding relationship query along with information for matching the corresponding relationship query to a received designated query.

14. The method of claim 1 wherein the rules are specified using a tag-based scripting language.

15. The method of claim 1, further comprising presenting the indications of the corresponding relationship searches in a ranked order.

16. The method of claim 1 wherein the executing the corresponding relationship searches is performed asynchronously.

17. The method of claim 1 wherein the determining the one or matching rules that correspond to the designated query further comprises:

determining a set of matching ontology paths that correspond to an indicated entity or entity type of the designated query.

18. The method of claim 1, further comprising:

when a suggested search tip is indicated for execution, executing the corresponding relationship search; and

presenting search results of the executed relationship search with indications of the components of the relationship query performed by the executed relationship search.

19. The method of claim 18 wherein the indications of the components in the presented results are used to facilitate teaching a query language.

20. A computer-readable medium containing instructions that, when executed, control a computer processor to provide search navigation tips by performing a method comprising:

receiving an indication of a designated query;

determining one or more matching rules that correspond to the designated query, each rule having associated relationship search specification that indicates a corresponding relationship search;

for each determined rule, determining and executing the corresponding relationship search against the corpus;

for one or more corresponding relationship searches that succeed, forwarding an indication of the corresponding relationship search as a suggested search tip.

21. The computer readable medium of claim 20 comprising a computer memory.

22. The computer-readable medium of claim 20 wherein the receiving the indication of the designated query further comprises receiving an indication of one or more components of a relationship query.

23. The computer-readable medium of claim 20 wherein the one or more components of a relationship query are indicated using IQL.

24. The computer-readable medium of claim 20 wherein the receiving the indication of the designated query further comprises receiving an indication of more or more keywords.

25. The computer-readable medium of claim 20 wherein the receiving the indication of the designated query comprises receiving an indication of one or more of an entity, an entity type, an action, or an action type.

26. The computer-readable medium of claim 20, the method further comprising:

for the one or more corresponding relationship searches that succeed,

evaluating results of the corresponding relationship search that succeeds to determine a plurality of specific relationship queries to serve as additional suggested searches; and

forwarding indications of the plurality of specific relationship queries as additional suggested search tips.

27. The computer-readable medium of claim 20, the method further comprising:

caching search results of one or more corresponding relationship searches that succeed; and

when a suggested searched tip is indicated for execution, returning the cached results of the relationship search that corresponds to the suggested search.

28. The computer-readable medium of claim 20 wherein the rules are specified in a file using a tag-based scripting language.

29. The computer-readable medium of claim 20, the method further comprising presenting the suggested search tips in a ranked order.

30. The computer-readable medium of claim 20 wherein the executing the corresponding relationship searches is performed asynchronously.

31. An relationship search navigation tip engine comprising:

a computer processor;

a data repository configured to maintain rules that map between one or more relationship query components and corresponding specifications of relationship queries;

a storage medium comprising a tip search manager that, when executed on the computer processor, is configured to parse the rules stored in the data repository to generate mappings for between potential input queries and corresponding relationship searches to be executed; and

a storage medium comprising a tip searcher that, when executed on the computer processor, is configured to receive an input query, determine one or more mappings that correspond to the input query, execute the relationship searches that correspond to the determined one or more mappings to generate corresponding search results, and return indications to the executed relationship searches that generated successful corresponding search results.

32. The navigation tip engine of claim 31 wherein the indications to the executed relationship searches that generated successful corresponding search results are visual navigation tips.

33. The navigation tip engine of claim 31 wherein the tip searcher is further configured to provide additional indications to relationship searches that correspond to specific entities or specific actions discovered during the executed relationship searches.

34. The navigation tip engine of claim 31 wherein the additional indications comprise deep tips.

35. The navigation tip engine of claim 31 wherein the tip searcher is configured to execute the relationship searches that correspond to the determined one or more mappings in parallel.

36. The navigation tip engine of claim 31 wherein the tip searcher is configured to execute the relationship searches that correspond to the determined one or more mappings asynchronously.

37. A computer-implemented method for performing a relationship search against a corpus of documents comprising:

executing a relationship search for an designated entity and/or designated action to yield a plurality of indications of documents;

determining whether each of the indicated documents contains a term that matches a designated ontology path specification; and

for each determined document, returning an indication of the document as a successful result of the relationship search.

38. A computer-implemented method for performing a relationship search against a corpus of documents comprising:

receiving an indication of a designated query;

executing a relationship search that corresponds to the designated query to yield one or more indications of documents, where each document contains at least one sentence or clause that matches the relationship search that corresponds to the designated query; and

determining whether a context surrounding the at least one matching sentence or clause further matches an entity or an action specified by the designated query.

39. The method of claim 38 wherein the context includes a sentence before the at least one matching sentence.

40. The method of claim 38 wherein the context includes a sentence after the at least one matching sentence.

41. The method of claim 38 wherein the designated query specifies at least one of a source entity, a target entity, or an action and specifies a context operator along with a corresponding entity or action.

42. A computer-implemented method of exporting result data from a relationship search performed against a corpus of documents, comprising:

executing a relationship search on the corpus of documents according to a relationship query to generate a search result having one or more sentences that match the query;

providing a visual interface for indicating one or more columns for formatting data from the search result and, for each such column, indicating a respective value that corresponds to a component of a relationship query; and

automatically formatting and storing the search result according to the one or more columns and their associated values by retrieving one or more terms from each sentence of the search result and storing the retrieved terms according to the indicated columns and their respective values.

43. The method of claim 42, the storing the retrieved terms according to the indicated columns and their respective values further comprising:

storing the retrieved terms according to the indicated columns and their respective values in a separator delimited file.

44. The method of claim 42 wherein the providing the visual interface further comprises:

providing a first area for selecting a number of columns, a second area for selecting a type of value for each column; and

a user interface control for selecting one of a source, a target, an action, an entity type, an action type, a grammatical role, or a type of metadata.

45. The method of claim 42, further comprising exporting the stored data as a data structure to be used with a statistical analysis tool.

46. The method of claim 45 wherein the statistical analysis tool is a spreadsheet program or an environment for executing a statistical language.

47. A relationship search facility for exporting result data from a relationship search performed against a corpus of documents, comprising:

a search engine that is configured to execute a relationship search on the corpus of documents according to a received relationship query to generate a search result having one or more sentences that match the received query;

a visual interface that is configured to, when executed, indicate one or more columns for formatting data from the search result and, for each such column, indicate a respective value that corresponds to a component of a relationship query; and

an export facility that is configured to, when executed, automatically format and store the search result according to the one or more columns and their associated values.