Text information analysis apparatus and method

• US 20020016798A1
• Filed: 07/18/2001
• Published: 02/07/2002
• Est. Priority Date: 07/25/2000
• Status: Active Grant

• Current Assignee: Kabushiki Kaisha Toshiba (Toshiba Corporation)
• Original Assignee: Kabushiki Kaisha Toshiba (Toshiba Corporation)
• Inventors: Suzuki, Masaru, Sakai, Tetsuya
• Granted Patent: US 7,099,819 B2
• Time in Patent Office: Days
• US Class Current: 715/517
• CPC Class Codes: G06F 16/353 into predefined classes
• CPC Class Codes: G06F 16/355 Class or cluster creation or modification
• CPC Class Codes: G06F 18/23 Clustering techniques
• CPC Class Codes: G06F 40/268 Morphological analysis
• CPC Class Codes: G06F 40/30 Semantic analysis

Abstract

Text information analysis apparatus arranges a plurality of texts according to the content of each text. In the text information analysis apparatus, a category decision unit classifies text to one of a plurality of predetermined categories. A cluster generation unit clusters texts having similar contents from the plurality of texts. A control unit controls said category decision unit and said cluster generation unit to simultaneously execute a category decision and clustering for the plurality of texts.

Claims

1. A text information analysis apparatus for arranging a plurality of texts according to the content of each text, comprising:

a category decision unit configured to classify text to one of a plurality of predetermined categories;

a cluster generation unit configured to cluster texts having similar contents from the plurality of texts; and

a control unit configured to control said category decision unit and said cluster generation unit to simultaneously execute a category decision and clustering for the plurality of texts.

2. The text information analysis apparatus according to claim 1,

further comprising an input unit configured to input the plurality of texts, and

wherein said control unit respectively supplies the plurality of texts input by said input unit to said category decision unit and to said cluster generation unit.

3. The text information analysis apparatus according to claim 1,

further comprising a memory configured to correspondingly store a category decision result by said category decision unit and a clustering result by said cluster generation unit, and

an output unit configured to comparatively output the category decision result and the clustering result from said memory through said control unit.

4. The text information analysis apparatus according to claim 3,

wherein said control unit supplies the category decision result from said category decision unit to said output unit before said cluster generation unit completes clustering the plurality of texts.

5. The text information analysis apparatus according to claim 3,

wherein said output unit discriminatively displays the texts classified to a predetermined category in the category decision result and the clustering result.

6. The text information analysis apparatus according to claim 2,

wherein said control unit supplies the plurality of texts to said category decision unit first, and then supplies the texts classified to each category to said cluster generation unit in case of receiving the category decision result from said category decision unit.

7. The text information analysis apparatus according to claim 2,

wherein said control unit supplies the plurality of texts to said category decision unit first, and then supplies the texts not classified to any category to said cluster generation unit in case of receiving the category decision result from said category decision unit.

8. The text information analysis apparatus according to claim 7,

in case of receiving the clustering result of the texts not classified to every categories from said cluster generation unit,

wherein said control unit selects a predetermined number of clusters in order of larger number of texts in each cluster from the clustering result, extracts feature words from the texts in the selected cluster, and stores a new category of which category decision rule includes the feature words of the selected cluster in said memory.

9. The text information analysis apparatus according to claim 3,

wherein said control unit extracts feature words from texts in each cluster of the clustering result, extracts a category decision rule of the category classifying the texts in the cluster from said memory, and stores the category decision rule including the feature words of corresponding cluster in said memory.

10. The text information analysis apparatus according to claim 6,

wherein said control unit extracts a category decision rule of each category and the clustering result of the texts classified to the category from said memory, extracts feature words from the texts of each cluster of the clustering result, and stores the category decision rule including the feature words of corresponding cluster in said memory.

11. The text information analysis apparatus according to claim 7,

wherein said control unit extracts each cluster of texts not classified to any category, a category decision rule of each category, and texts classified to the category from said memory;

extracts feature words from the texts of the cluster and feature words from the texts classified to the category; and

stores the category decision rule including the feature words of the texts of the cluster in said memory if the feature words of the texts of the cluster are similar to the feature words of the texts classified to the category.

13. The text information analysis method according to claim 12, further comprising:

correspondingly storing a category decision result and a clustering result in a memory, and

comparatively outputting the category decision result and the clustering result through a display.

14. The text information analysis method according to claim 13,

at the outputting step, further comprising:

outputting the category decision result while simultaneously executing clustering of the plurality of texts.

15. The text information analysis method according to claim 13,

at the outputting step, further comprising:

discriminatively displaying the texts classified to a predetermined category in the category decision result and the clustering result.

16. The text information analysis method according to claim 12,

at the clustering step, further comprising:

clustering the texts classified to each category by unit of category.

17. The text information analysis method according to claim 12,

at the clustering step, further comprising:

clustering the texts not classified to any category.

18. The text information analysis method according to claim 17,

at the controlling step, further comprising:

selecting a predetermined number of clusters in order of larger number of texts in each cluster from the clustering result;

extracting feature words from the texts in the selected cluster; and

storing a new category of which category decision rule includes the feature words of the selected cluster in said memory.

19. The text information analysis method according to claim 13,

at the controlling step, further comprising:

extracting feature words from texts in each cluster of the clustering result;

extracting a category decision rule of the category classifying the texts in the cluster from said memory; and

storing the category decision rule including the feature words of corresponding cluster in said memory.

20. The text information analysis method according to claim 16,

at the controlling step, further comprising:

extracting a category decision rule of each category and the clustering result of the texts classified to the category from said memory;

extracting feature words from the texts of each cluster of the clustering result; and

storing the category decision rule including the feature words of the corresponding cluster in said memory.

21. The text information analysis method according to claim 17,

at the controlling step, further comprising:

extracting each cluster of texts not classified to any category, a category decision rule of each category, and texts classified to the category from said memory;

extracting feature words from the texts of the cluster and feature words from the texts classified to the category; and

storing the category decision rule including the feature words of the texts of the cluster in said memory if the feature words of the texts of the cluster are similar to the feature words of the texts classified to the category.

12. A text information analysis method for arranging a plurality of texts according to the content of each text, comprising:

classifying text to one of a plurality of predetermined categories;

clustering texts having similar contents from the plurality of texts; and

simultaneously executing a category decision and clustering for the plurality of texts.

22. A computer program product for use with a computer, said computer program product comprising:

a computer usable medium having computer readable program code embodied in said medium for causing said computer to arrange a plurality of texts according to the content of each text, said computer readable program code having;

computer readable program code to classify text to one of a plurality of predetermined categories;

computer readable program code to cluster texts having similar contents from the plurality of texts; and

computer readable program code to simultaneously execute a category decision and clustering for the plurality of texts.