Text processing and retrieval system and method

• US 5,963,965 A
• Filed: 02/18/1997
• Issued: 10/05/1999
• Est. Priority Date: 02/18/1997
• Status: Expired due to Term

• Current Assignee: Amobee, Inc. (Nexxen International Ltd.)
• Original Assignee: Semio (Nexxen International Ltd.)
• Inventors: Vogel, Claude
• Primary Examiner(s): Hong, Stephen S.
• Application Number:
  US08/801,970
• Time in Patent Office: 959 Days
• Field of Search: 707/501, 707/514, 707/531, 707/3, 707/4, 707/5, 707/104, 704/1, 704/9
• US Class Current: 715/236
• CPC Class Codes: G06F 16/338 Presentation of query results
• CPC Class Codes: Y10S 707/99933 Query processing, i.e. searching

Abstract

A content-based system and method for text processing and retrieval is provided wherein a plurality of pieces of text are processed based on content to generate an index for each piece of text, the index comprising a list of phrases that represent the content of the piece of text. The phrases are grouped together to generate clusters based on a degree of relationship of the phrases, and a hierarchical structure is generated, the hierarchical structure comprising a plurality of maps, each map corresponding to a predetermined degree of relationship, the map graphically depicting the clusters at the predetermined degree of relationship, and comprising a plurality of nodes, each node representing a cluster, and a plurality of links connecting nodes that are related. The map is displayed to a user, a user selects a particular cluster on the map, and a portion of text is extracted from said pieces of text based on the cluster selected by the user. The system may also generate scenarios, based on said maps, that indicate changes in the relationships shown by the maps.

Claims

1. A content-based text processing and retrieval system, comprising:

means for processing a plurality of pieces of text based on content to generate an index for each piece of text, the index comprising a list of phrases that represent the content of the piece of text;

means for grouping phrases together to generate clusters based on a predetermined degree of relationship between the phrases;

means for generating a hierarchical structure, the hierarchical structure comprising a plurality of maps, each map corresponding to a predetermined degree of relationship, the map graphically depicting the clusters at the predetermined degree of relationship and comprising a plurality of nodes, each node representing a cluster, and a plurality of links connecting nodes that are related;

means for selecting a predetermined map;

means for displaying said selected map to a user;

means for selecting a particular cluster displayed on said selected map; and

means for extracting a portion of text from said pieces of text based on the selected cluster.

2. The system of claim 1, wherein said processing means comprises means for gathering a plurality of pieces of text, means for extracting a lexicon from said gathered pieces of text, the lexicon comprising a list of phrases that indicate the content of said pieces of text, and means for comparing said lexicon to each piece of text to generate an index for each piece of text.

3. The system of claim 2, wherein said gathering means comprises a software application for gathering pieces of text from the Internet.

4. The system of claim 2, wherein said lexicon extracting means comprises means for removing empty words from said pieces of text, means for extracting proper and common nouns from said pieces of text, means for extracting phrases that are repeated within a piece of text, and means for extracting noun phrases from said pieces of text.

5. The system of claim 4, wherein said means for extracting noun phrases comprises means for comparing a plurality of phrases within said pieces of text to a template in order to extract phrases having nouns.

6. The system of claim 2 further comprising means for generating a semiotic data structure based on said lexicon, said semiotic data structure comprising a tag that is associated with each word in said lexicon to classify which word by its content, and means for comparing a plurality of maps to each other based on the semiotic data structure to generate a scenario, said scenario indicating changes in the relationships graphically depicted by said maps.

7. The system of claim 6, wherein said tag is selected from one of a tag indicating a person, a tag indicating a function and a tag indicating a topic.

8. The system of claim 2 further comprising means for filtering said indexes based on a filter criteria selected by a user to select a predetermined number of pieces of text, and means for generating one or more word clusters from the indexes of said predetermined number of pieces of text.

9. A method for content-based text processing and retrieval, comprising:

processing a plurality of pieces of text based on content to generate an index for each piece of text, the index comprising a list of phrases that represent the content of the piece of text;

grouping phrases together to generate clusters based on a predetermined degree of relationship between the phrases;

generating a hierarchical structure, the hierarchical structure comprising a plurality of maps, each map corresponding to a predetermined degree of relationship, the map graphically depicting the clusters at the predetermined degree of relationship and comprising a plurality of nodes, each node representing a cluster, and a plurality of links connecting nodes that are related;

selecting a predetermined map;

displaying said selected map to a user;

selecting a particular cluster displayed on said selected map; and

extracting a portion of text from said pieces of text based on the selected cluster.

10. The method of claim 9, wherein processing comprises gathering a plurality of pieces of text, extracting a lexicon from said gathered pieces of text, the lexicon comprising a list of phrases that indicate the content of said pieces of text, and comparing said lexicon to each piece of text to generate an index for each piece of text.

11. The method of claim 10, wherein gathering comprises using a software application to gather pieces of text from the Internet.

12. The method of claim 10, wherein extracting said lexicon comprises removing empty words from said pieces of text, extracting proper and common nouns from said pieces of text, extracting phrases that are repeated within a piece of text, and extracting noun phrases from said pieces of text.

13. The method of claim 12, wherein extracting said noun phrases comprises comparing a plurality of phrases within said pieces of text to a template in order to extract phrases having nouns.

14. The method of claim 10 further comprising generating a semiotic data structure based on the lexicon, said semiotic data structure comprising a tag that is associated with each word in said semiotic data structure to classify each word by its content, and comparing a plurality of maps to each other, based on the semiotic data structure, to generate a scenario, said scenario indicating changes in the relationships shown by said maps.

15. The method of claim 14, wherein said tag is selected from one of a tag indicating a person, a tag indicating a function and a tag indicating a topic.

16. The method of claim 10 further comprising filtering said indexes based on a filter criteria selected by a user to select a predetermined number of pieces of text, and generating one or more word clusters from the indexes of said predetermined number of pieces of text.

17. A content-based text processing and retrieval system, comprising:

means for processing a plurality of pieces of text based on content to generate an index for each piece of text, the index comprising a list of phrases that represent the content of the piece of text;

means for grouping phrases together to generate clusters based on a predetermined degree of relationship between the phrases; and

means for generating a hierarchical structure, the hierarchical structure comprising a plurality of maps, each map corresponding to a predetermined degree of relationship, the map graphically depicting the clusters at the predetermined degree of relationship and comprising a plurality of nodes, each node representing a cluster, and a plurality of links connecting nodes that are related.

18. The system of claim 17, wherein said processing means comprises means for gathering a plurality of pieces of text, means for extracting a lexicon from said gathered pieces of text, the lexicon comprising a list of phrases that indicate the content of said pieces of text, and means for comparing said lexicon to each piece of text to generate an index for each piece of text.

19. The system of claim 18, wherein said gathering means comprises a software application for gathering pieces of text from the Internet.

20. The system of claim 18, wherein said lexicon extracting means comprises means for removing empty words from said pieces of text, means for extracting proper and common nouns from said pieces of text, means for extracting phrases that are repeated within a piece of text, and means for extracting noun phrases from said pieces of text.

21. The system of claim 20, wherein said means for extracting noun phrases comprises means for comparing a plurality of phrases within said pieces of text to a template in order to extract phrases having nouns.

22. The system of claim 18 further comprising means for generating a semiotic data structure based on said lexicon, said semiotic data structure comprising a tag that is associated with each word in said lexicon to classify which word by its content, and means for comparing a plurality of maps to each other, based on said semiotic data structure, to generate a scenario, said scenario indicating changes in the relationships graphically depicted by said maps.

23. The system of claim 22, wherein said tag is selected from one of a tag indicating a person, a tag indicating a function and a tag indicating a topic.

24. The system of claim 18 further comprising means for filtering said indexes based on a filter criteria selected by a user to select a predetermined number of pieces of text, and means for generating one or more word clusters from the indexes of said predetermined number of pieces of text.

25. A method for content-based text processing and retrieval system, comprising:

processing a plurality of pieces of text based on content to generate an index for each piece of text, the index comprising a list of phrases that represent the content of the piece of text;

grouping phrases together to generate clusters based on a predetermined degree of relationship between the phrases; and

generating a hierarchical structure, the hierarchical structure comprising a plurality of maps, each map corresponding to a predetermined degree of relationship, the map graphically depicting the clusters at the predetermined degree of relationship and comprising a plurality of nodes, each node representing a cluster, and a plurality of links connecting nodes that are related.

26. The method of claim 25, wherein processing comprises gathering a plurality of pieces of text, extracting a lexicon from said gathered pieces of text, the lexicon comprising a list of phrases that indicate the content of said pieces of text, and comparing said lexicon to each piece of text to generate an index for each piece of text.

27. The method of claim 26, wherein gathering comprises using a software application to gather pieces of text from the Internet.

28. The method of claim 26, wherein extracting said lexicon comprises removing empty words from said pieces of text, extracting proper and common nouns from said pieces of text, extracting phrases that are repeated within a piece of text, and extracting noun phrases from said pieces of text.

29. The method of claim 28, wherein extracting said noun phrases comprises comparing a plurality of phrases within said pieces of text to a template in order to extract phrases having nouns.

30. The method of claim 26 further comprising generating a semiotic data structure based on the lexicon, said semiotic data structure comprises a tag that is associated with each word in said semiotic data structure to classify each word by its content, and comparing a plurality of maps to each other, based on said semiotic data structure, to generate a scenario, said scenario indicating changes in the relationships shown by said maps.

31. The method of claim 30, wherein said tag is selected from one of a tag indicating a person, a tag indicating a function and a tag indicating a topic.

32. The method of claim 26 further comprising filtering said indexes based on a filter criteria selected by a user to select a predetermined number of pieces of text, and generating one or more word clusters from the indexes of said predetermined number of pieces of text.

33. A content-based text processing and retrieval system, comprising:

means for processing a plurality of pieces of text based on content to generate an index for each piece of text, the index comprising a list of phrases that represent the content of the piece of text;

means for grouping phrases together to generate clusters based on a predetermined degree of relationship between the phrases;

means for generating a hierarchical structure, the hierarchical structure comprising a plurality of maps, each map corresponding to a predetermined degree of relationship, the map graphically depicting the clusters at the predetermined degree of relationship and comprising a plurality of nodes, each node representing a cluster, and a plurality of links connecting nodes that are related;

means for generating a semiotic data structure from said plurality of pieces of text, the semiotic data structure comprising a list of phrases that indicate the content of said pieces of text, and a tag that is associated with each phrase in said semiotic data structure to classify which word by its content; and

means for comparing a plurality of maps to each other to generate a scenario, said scenario indicating changes in the relationship graphically depicted by said maps.

34. The system of claim 33, wherein said tag is selected from one of a tag indicating a person, a tag indicating a function and a tag indicating a topic.