System for Automatically Generating Queries

• US 20100250547A1
• Filed: 06/10/2010
• Published: 09/30/2010
• Est. Priority Date: 08/13/2001
• Status: Active Grant

• Current Assignee: III Holdings 6, LLC (Intellectual Ventures LLC)
• Original Assignee: Xerox Corporation (Xerox Holdings Corp.)
• Inventors: Shanahan, James G., Grefenstette, Gregory T.
• Granted Patent: US 8,219,557 B2
• Time in Patent Office: Days
• US Class Current: 707/740
• CPC Class Codes: G06F 16/30 of unstructured textual data document management systems G06F16/93
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation

Abstract

A method, system and article of manufacture therefor, are disclosed for automatically generating a query from document content.

Claims

1. A computer implemented method, comprising:

recording in a database entities having associated therewith an entity-type identifying whether the entity belongs to one or more predefined classes of entities;

automatically recognizing in selected content at least one entity recorded in the database of entities;

automatically assigning the selected content a set of one or more categories using a categorization system for categorizing content;

the categorization system having categories corresponding to categories of content available through an information retrieval system;

automatically acquiring search results by querying content available through the information retrieval system using the at least one entity recognized in the selected content;

said acquiring contextualizing said querying using the set of one or more categories assigned to the selected content;

automatically using the search results acquired through the information retrieval system for enriching the selected content;

making available the selected content enriched using the search results.

2. The method according to claim 1, further comprising automatically filtering the search results before enriching the selected content.

3. The method according to claim 1, wherein said recognizing recognizes the at least one entity from proper names, times, locations, amounts, citations, and addresses.

4. The method according to claim 1, wherein said recognizing recognizes the at least one entity as a proper name.

5. The method according to claim 1, wherein said recognizing recognizes the at least one entity as a location.

6. The method according to claim 1, wherein said recognizing recognizes the at least one entity using one or a combination of regular expressions, lexicons, keywords, and rules.

7. The method according to claim 1, wherein entities recorded in the database further comprise lexicons that have associated therewith an entity and an entity-type.

8. The method according to claim 7, wherein the lexicons recorded in the database have further associated therewith an entity-string.

9. The method according to claim 8, wherein the lexicons recorded in the database have further associated therewith a part-of-speech tag.

10. The method according to claim 1, wherein said recognizing further comprises tagging the at least one entity in the selected document content with a part-of-speech tag that identifies grammatical usage of the at least one entity.

11. The method according to claim 1, wherein said categorizing the selected document content further comprises assigning the selected document content a category vocabulary associated with each category in the set of one or more categories.

12. The method according to claim 11, wherein said acquiring further contextualizes said querying using the category vocabulary associated with each category in the set of one or more categories.

13. The method according to claim 12, wherein said acquiring further contextualizes said querying using one or more aspects to augment the at least one entity recognized in the selected-document content.

14. The method according to claim 1, wherein the categories of the categorization system are organized as a hierarchy.

15. The method according to claim 1, wherein said assigning the selected content a set of one or more categories further comprises computing a similarity measure between the selected content and class profiles of categories in the categorization system.

16. The method according to claim 1, wherein the search results used to enrich the selected content are categorized in a category that is not in the set of one or more categories.

17. The method according to claim 16, wherein the category that is not in the set of one or more categories is a parent category of a category in the set of one or more categories.

18. The method according to claim 1, wherein said acquiring further contextualizes said querying using one or more aspects to augment the at least one entity recognized in the selected content.

19. A computer implemented method for automatically generating a query from document content to search for information using an information retrieval system, comprising:

defining an ontology organized by categories of information;

automatically identifying, using a first processor, a set of entities in the document content;

automatically categorizing, using the first or a second processor, the document content by assigning the document content at least one category of information from the ontology; and

automatically formulating the query for searching, using the information retrieval system, for information related to the set of entities and the at least one assigned category of information.

20. The method of claim 19, wherein entities in the set of entities are weighted entities.

21. The method of claim 19, further comprising directing the query to the at least one assigned category of information.

22. The method of claim 19, wherein the document content is formatted as HTML.

23. The method of claim 19, wherein the document content is from a document suitable for viewing on a browser.

24. The method of claim 19, wherein the categories of information in the ontology are organized hierarchically.

25. The method of claim 19, wherein said automatically categorizing the document content is performed using a learning module.

26. The method of claim 19, wherein said automatically categorizing the document content is performed using at least one of a probabilistic model, a Bayesian model, a fuzzy model, a Boolean-valued logic model, a nearest-neighbor approach, a neural network, a latent semantic indexing model, and a vector space model.

27. A computer implemented method for automatically generating a query from document content to search for information using an information retrieval system utilizing an ontology organized by categories of information, comprising:

automatically identifying, using a first processor, a set of entities in the document content;

automatically categorizing, using the first or a second processor, the document content by assigning the document content at least one category of information from the ontology; and

automatically formulating the query for searching, using the information retrieval system, for information related to the set of entities and the at least one assigned category of information.

28. The method of claim 27, wherein entities in the set of entities are weighted entities.

29. The method of claim 27, further comprising directing the query to the at least one assigned category of information.

30. The method of claim 27, wherein the document content is formatted as HTML.

31. The method of claim 27, wherein the document content is from a document suitable for viewing on a browser.

32. The method of claim 27, wherein the categories of information in the ontology are organized hierarchically.

33. The method of claim 27, wherein said automatically categorizing the document content is performed using a learning module.

34. The method of claim 27, wherein said automatically categorizing the document content is performed using at least one of a probabilistic model, a Bayesian model, a fuzzy model, a Boolean-valued logic model, a nearest-neighbor approach, a neural network, a latent semantic indexing model, and a vector space model.

35. A computer implemented method for retrieving information, comprising:

automatically identifying, using a first processor, a set of entities in document content;

automatically categorizing, using the first or a second processor, the document content by assigning to the document content at least one category of information from an ontology organized by categories of information;

automatically formulating a query for searching for information related to the set of entities and the at least one assigned category of information; and

automatically retrieving the information using the query.

36. The method of claim 35, wherein entities in the set of entities are weighted entities.

37. The method of claim 35, further comprising directing the query to the at least one assigned category of information.

38. The method of claim 35, wherein the document content is formatted as HTML.

39. The method of claim 35, wherein the document content is from a document suitable for viewing on a browser.

40. The method of claim 35, wherein the categories of information in the ontology are organized hierarchically.

41. The method of claim 35, wherein said automatically categorizing the document content is performed using a learning module.

42. The method of claim 35, wherein said automatically categorizing the document content is performed using at least one of a probabilistic model, a Bayesian model, a fuzzy model, a Boolean-valued logic model, a nearest-neighbor approach, a neural network, a latent semantic indexing model, and a vector space model.

43. The method of claim 35, further comprising automatically inserting the retrieved information in a document that contains the document content.

44. The method of claim 43, wherein said inserting inserts the retrieved information as a hyperlink.

45. The method of claim 35, further comprising enriching the document content using the retrieved information.

46. The method of claim 45, wherein said enriching uses the retrieved information to enrich the document content with a hyperlink.