Methods and systems for detecting and extracting information

• US 7,689,536 B1
• Filed: 12/18/2003
• Issued: 03/30/2010
• Est. Priority Date: 12/18/2003
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Weissman, Adam J., Elbaz, Gilad Israel
• Primary Examiner(s): Fleurantin; Jean B.
• Assistant Examiner(s): MYINT, DENNIS Y
• Application Number:
  US10/741,303
• Time in Patent Office: 2,294 Days
• Field of Search: 395/605, 707/1, 707/2, 707/100, 707/3, 434/118
• US Class Current: 1/1
• CPC Class Codes: G06F 16/00 Information retrieval; Database structures therefor; File system structures therefor
• CPC Class Codes: Y10S 707/99932 Access augmentation or optimizing

Abstract

Systems and methods that detect information and extract information are described. In one aspect, target rules are defined for detection of target hits in an article, including defining a target article region, extraction rules are defined based on the target rules for the extraction of extracts from the article, including an extraction article region, target rules are applied to each target article region of the article to determine target hits, and extraction rules are applied to detect at least one extract from the article based on the determined target hit.

Claims

1. A method for extracting information from an electronic document using a computer server system having one or more processors, the method comprising:

receiving, at the server system, a request for information that includes a definition of a concept list comprising an origin concept, a relationship between the origin concept and an evaluated concept, and a distance representing a strength of the relationship between the origin concept and the evaluated concept, and a target scope that characterizes a size of document regions to which the concept list is to be applied, wherein the request for information and the target scope are received from a user interacting with the server system through a client device connected to the server system via a network;

receiving, at the server system a definition of an extraction rule, wherein the extraction rule definition comprises an extraction scope that characterizes document regions to be extracted;

the server system determining a target score for document regions of the article, wherein the score represents how well document regions of the size to which the concept list is to be applied relate to the concept list;

the server system applying the extraction rule to the article to extract document regions characterized by the extraction scope from the article, wherein the application of the extraction rule is based on the determined target score; and

outputting the extract from the server system for the client device in response to the request for information.

2. The method of claim 1, wherein applying the extraction rule comprises extracting a plurality of extracts.

3. The method of claim 2, further comprising the server system sorting the extracts based on the extraction rule.

4. The method of claim 1, further comprising the server system selecting a first extract from the article for output based on the target score.

5. The method of claim 1, further comprising receiving, at the server system, a target score formula for determining the target score.

6. The method of claim 5, wherein determining the target score comprises using the target score formula.

7. The method of claim 1, wherein the request for information further comprises a gist defined as a vector of weighted concepts.

8. The method of claim 7, wherein the gist comprises a user defined gist.

9. The method of claim 7, wherein the gist comprises a calculated gist of a document region.

10. The method of claim 1, wherein the request for information further comprises a concept set that comprises the concept list and a second concept.

11. The method of claim 10, wherein the second concept comprises a product of set operations on two or more other concepts.

12. The method of claim 10, wherein the concept set further comprises at least one set operation.

13. The method of claim 12, wherein the set operation comprises one of “

AND”

, “

OR”

, and “

AND NOT”

.

14. The method of claim 1, wherein the document region characterized by the target scope comprises an article, a sentence, or a term.

15. The method of claim 1, wherein the document region characterized by the extraction scope comprises an article, a sentence, or a term.

16. The method of claim 1, further comprising the server system preprocessing the article, wherein preprocessing comprises:

determining at least one concept associated with the article; and

determining a gist of the article.

17. The method of claim 1, wherein

the origin concept comprises a group of related words, relationships with other concepts, the strengths of the relationships, and statistics regarding the usage of the origin concept in a language.

18. The method of claim 1, wherein receiving the definition of the extraction rule further comprises receiving a definition of a sort order in which extracts are to be sorted for output.

19. The method of claim 1, wherein the distance comprises a numeric representation of the strength of the relationship between the origin concept and the evaluated concept.

20. The method of claim 1, wherein the relationship comprises one of “

is a product of”

, “

is a part of”

, “

is a kind of”

, “

has kind”

, or “

has part”

.

21. The method of claim 1, wherein the origin concept comprises at least one search term.

22. An article for extracting information from an electronic document using a computer server system having one or more processors, the article comprising one or more computer-readable data storage media containing program code operable to cause one or more machines of a server system to perform operations, the operations comprising:

receiving, from a user, a request for information that includes a definition of a concept list comprising an origin concept, a relationship between the origin concept and an evaluated concept, and a distance representing a strength of the relationship between the origin concept and the evaluated concept, and a target scope that characterizes a size of document regions to which the target rule is to be applied;

receiving, from the user, a definition of an extraction rule, wherein the extraction rule definition comprises an extraction scope that characterizes document regions to be extracted;

determining a target score for document regions of the article, wherein the score represents how well document regions of the size to which the concept list is to be applied relate to the concept list;

applying the extraction rule to the article to extract document regions characterized by the extraction scope from the article, wherein the application of the extraction rule is based on the determined target score; and

outputting the extract in response to the request for information for use by a client device.

23. The article of claim 22, wherein applying the extraction rule comprises extracting a plurality of extracts.

24. The article of claim 23, further comprising sorting the extracts based on the extraction rule.

25. The article of claim 22, further comprising selecting a first extract from the article for output based on the target score.

26. The article of claim 22, further comprising receiving a target score formula for determining the target score.

27. The article of claim 26, wherein determining the target score comprises using the target score formula.

28. The article of claim 22, wherein the request for information further comprises a gist defined as a vector of weighted concepts.

29. The article of claim 28, wherein the gist comprises a user defined gist.

30. The article of claim 28 wherein the gist comprises a calculated gist of a document region.

31. The article of claim 22, wherein the request for information further comprises a concept set that comprises the concept list and a second concept.

32. The article of claim 31, wherein the second concept comprises a product of set operations on two or more other concepts.

33. The article of claim 31, wherein the concept set further comprises at least one set operation.

34. The article of claim 33, wherein the set operation comprises one of “

AND”

, “

OR”

, and “

AND NOT”

.

35. The article of claim 22, wherein the document region characterized by the target scope comprises an article, a sentence, or a term.

36. The article of claim 22, wherein the document region characterized by the extraction scope comprises an article, a sentence, or a term.

37. The article of claim 22, further comprising preprocessing the article, wherein preprocessing comprises:

determining at least one concept associated with the article; and

determining a gist of the article.

38. The article of claim 22, wherein receiving the definition of the extraction rule further comprises receiving a definition of a sort order in which extracts are to be sorted for output.

39. The article of claim 22, wherein the distance comprises a numeric representation of the strength of the relationship between the origin concept and the evaluated concept.

40. The article of claim 22, wherein the relationship comprises one of “

is a product of”

, “

is a part of”

, “

is a kind of”

, “

has kind”

, or “

has part”

.

41. The article of claim 22, wherein the origin concept comprises at least one search term.

42. A computer-implemented method for extracting a subset of a document using a computer server system having one or more processors, the method comprising:

receiving, at the server system, a request for information that describes a combination of two or more concept lists, wherein each concept list is defined by an origin concept, a relationship between the origin concept and an evaluated concept, and a distance representing a strength of the relationship between the origin concept and the evaluated concept, wherein the two or more concept lists are combined using an operation to define a target definition that is to be detected, wherein the request for information is received from a user interacting with the server system through a client device connected to the server system via a network;

receiving, at the server system, a description of a document region targeted for extraction;

accessing a document stored in a document database using the server system;

based on the target definition and the document regions targeted for extraction, using the server system to extract one or more regions of the accessed document; and

the server system outputting the extracted regions for the client device in response to the request for information.

43. The method of claim 42, wherein the origin concepts each comprise a lexical concept defined by a group of related words and relationships with related concepts.