Document clustering methods, document cluster label disambiguation methods, document clustering apparatuses, and articles of manufacture

• US 7,636,730 B2
• Filed: 04/29/2005
• Issued: 12/22/2009
• Est. Priority Date: 04/29/2005
• Status: Expired due to Fees

• Current Assignee: Battelle Memorial Institute
• Original Assignee: Battelle Memorial Research
• Inventors: Calapristi, Augustin J., Sanfilippo, Antonio, Crow, Vernon L., Turner, Alan E., Hetzler, Elizabeth G.
• Primary Examiner(s): Jalil; Neveen Abel
• Assistant Examiner(s): Hicks; Michael J
• Application Number:
  US11/118,156
• Publication Number:
  US 20060248053A1
• Time in Patent Office: 1,698 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06F 16/355 Class or cluster creation or modification
• CPC Class Codes: Y10S 707/99942 Manipulating data structure, e.g. compression, compaction, compilation
• CPC Class Codes: Y10S 707/99943 Generating database or data structure, e.g. via user interface

Abstract

Document clustering methods, document cluster label disambiguation methods, document clustering apparatuses, and articles of manufacture are described. In one aspect, a document clustering method includes providing a document set comprising a plurality of documents, providing a cluster comprising a subset of the documents of the document set, using a plurality of terms of the documents, providing a cluster label indicative of subject matter content of the documents of the cluster, wherein the cluster label comprises a plurality of word senses, and selecting one of the word senses of the cluster label.

Claims

1. A document clustering method comprising:

providing a document set comprising a plurality of documents;

providing a cluster comprising a subset of the documents of the document set, wherein the subset comprises a plurality of the documents;

using a plurality of terms of the documents of the cluster which are indicative of subject matter content of the documents of the cluster, selecting a cluster label indicative of the subject matter content of the documents of the cluster, wherein the cluster label is selected at least in part by co-occurrence of the cluster label and the plurality of terms of the documents of the cluster and wherein the cluster label comprises a plurality of word senses; and

selecting one of the word senses of the cluster label having an increased relevancy with respect to the plurality of terms of the documents of the cluster compared with the relevancies of others of the word senses.

2. The method of claim 1 wherein the providing the cluster comprises identifying the documents of the subset using terms occurring in the documents of the subset.

3. The method of claim 2 wherein the selecting the cluster label comprises selecting one of the terms.

4. The method of claim 3 further comprising analyzing the documents of the document set to identify the terms.

5. The method of claim 3 wherein the terms comprise minor terms and major terms, and the cluster label comprises a major term.

6. The method of claim 1 wherein the selecting the one of the word senses comprises determining, for individual ones of the word senses, a plurality of semantic similarity values with respect to the terms, wherein the semantic similarity values are individually indicative of a degree of semantic similarity between one of the word senses and one of the terms.

7. The method of claim 6 wherein the selecting the one of the word senses comprises selecting the one of the word senses having a greatest cumulative semantic similarity with respect to the terms.

8. The method of claim 1 wherein the providing the cluster comprises identifying the documents of the subset using the terms occurring in the documents of the subset and wherein the selecting the one of the word senses comprises determining, for individual ones of the word senses, a plurality of semantic similarity values with respect to individual ones of the terms.

9. The method of claim 8 wherein the selecting the one of the word senses further comprises, for individual ones of the terms, selecting the greatest semantic similarity values for individual ones of the word senses and, for individual ones of the word senses, summing the selected greatest semantic similarity values for the respective word sense providing a cumulative value for the word sense, and selecting the word sense having the greatest cumulative value.

10. The method of claim 9 wherein the determining comprises determining, for individual ones of the word senses, the semantic similarity values with respect to a plurality of word senses of individual ones of the terms of the documents of the cluster.

11. The method of claim 1 further comprising translating the cluster label from a first language to a second language using the selected one of the word senses.

13. The method of claim 1 wherein the selecting the cluster label comprises selecting the cluster label after the providing the cluster, and wherein the selecting the one of the word senses comprises selecting the one of word senses of the cluster label after the selecting the cluster label comprising the plurality of word senses.

14. The method of claim 1 wherein the documents of the cluster individually comprise a plurality of the terms which individually comprise a word.

12. A document cluster label disambiguation method comprising:

selecting a cluster label for a cluster comprising a subset of a plurality of documents of a document set at least in part by co-occurrence of the cluster label and a plurality of terms of the documents of the cluster which are indicative of subject matter content of the documents of the cluster, wherein the subset comprises a plurality of the documents and wherein the cluster label comprises one of a plurality of terms common to at least some of the documents of the cluster and the cluster label comprises a plurality of word senses;

determining, for individual ones of the word senses, a plurality of semantic similarity values for respective ones of the terms, wherein the semantic similarity values are individually indicative of a degree of semantic similarity between one of the word senses and one of the terms;

analyzing the semantic similarity values determined for respective ones of the word senses;

selecting one of the word senses using the analyzing; and

wherein the selecting the one of the word senses comprises, using the terms of the documents of the cluster, selecting the one of the word senses having an increased relevancy with respect to the subject matter content of the documents of the cluster compared with the relevancies of others of the word senses.

15. The method of claim 12 wherein the determining comprises determining, for individual ones of the word senses, a plurality of semantic similarity values with respect to individual ones of the terms.

16. The method of claim 12 wherein the selecting the one of the word senses comprises selecting the one of the word senses having a greatest cumulative semantic similarity with respect to the terms.

17. The method of claim 12 wherein the analyzing comprises, for individual ones of the terms, selecting the greatest semantic similarity values for individual ones of the word senses and, for individual ones of the word senses, summing the selected greatest semantic similarity values for the respective word sense providing a cumulative value for the word sense;

and wherein the selecting the one of the word senses comprises selecting the one of the word senses having the greatest cumulative value.

18. The method of claim 12 wherein the terms comprise minor terms and major terms, and the cluster label comprises a major term.

19. The method of claim 12 wherein the providing selecting the cluster label for the cluster comprises selecting the cluster label after identifying the documents of the subset which correspond to the cluster, and wherein the selecting the one of the word senses comprises selecting the one of the word senses after the identifying the documents of the subset which correspond to the cluster.

20. A document clustering apparatus comprising:

processing circuitry configured to access a document set comprising a plurality of documents, to define a cluster comprising a subset of the documents of the document set and wherein the subset comprises a plurality of the documents, to identify a cluster label indicative of subject matter content of at least one of the documents of the cluster at least in part by co-occurrence of the cluster label and a plurality of terms of the documents of the cluster which are indicative of the subject matter content of the documents of the cluster, and to use the terms of the documents of the cluster to disambiguate the cluster label after the identification of the cluster label to increase the relevancy of the cluster label with respect to the subject matter content of the at least one of the documents compared with the cluster label prior to the disambiguation.

21. The apparatus of claim 20 wherein the cluster label comprises a plurality of word senses and the processing circuitry is configured to select one of the word senses to disambiguate the cluster label.

22. The apparatus of claim 21 wherein the processing circuitry is configured to select the one of the word senses having increased relevancy with respect to the subject matter content of the documents of the cluster compared with the relevancies of others of the word senses.

23. The apparatus of claim 21 wherein the processing circuitry is configured to determine, for individual ones of the word senses, a plurality of semantic similarity values with respect to the terms, wherein the semantic similarity values are individually indicative of a degree of semantic similarity between one of the word senses and one of the terms.

24. The apparatus of claim 23 wherein the processing circuitry is configured to select the one of the word senses having a greatest cumulative semantic similarity with respect to the terms.

25. The apparatus of claim 21 wherein the terms comprise minor terms and major terms, and the cluster label comprises a major term.

26. The apparatus of claim 21 wherein the processing circuitry is configured to identify the documents of the subset using the terms occurring in the documents of the subset, and wherein the processing circuitry is configured to determine, for individual ones of the word senses, a plurality of semantic similarity values with respect to the terms to select the one of the word senses.

27. The apparatus of claim 26 wherein the processing circuitry is configured, for individual ones of the terms, to select the greatest semantic similarity values for individual ones of the word senses and, for individual ones of the word senses, to sum the selected greatest semantic similarity values for the respective word sense providing a cumulative value for the word sense, and to select the word sense having the greatest cumulative value to select the one of the word senses.

28. The apparatus of claim 27 wherein the processing circuitry is configured to determine, for individual ones of the word senses, the semantic similarity values with respect to a plurality of word senses of individual ones of the terms.

29. The apparatus of claim 20 wherein the document set is external of the apparatus, and further comprising a communications interface configured to access the document set.

30. The apparatus of claim 20 further comprising a display, and wherein the processing circuitry is configured to control the display to depict the cluster and the selected one of the word senses of the cluster label.

31. The apparatus of claim 20 wherein the processing circuitry is configured to identify the cluster label after the defining of the cluster.

32. The apparatus of claim 20 wherein the documents of the cluster individually comprise a plurality of words.

33. An article of manufacture comprising:

a computer-readable storage medium comprising programming configured to cause processing circuitry to;

select a cluster label for a cluster comprising a subset of a plurality of documents of a document set at least in part by co-occurrence of the cluster label and a plurality of terms of the documents of the cluster which are indicative of subject matter content of the documents of the cluster, wherein the subset comprises a plurality of the documents and wherein the cluster label comprises one of a plurality of terms common to at least one of the documents of the cluster and the cluster label comprises a plurality of word senses;

determine, for individual ones of the word senses, a plurality of semantic similarity values for respective ones of the terms, wherein the semantic similarity values are individually indicative of a degree of semantic similarity between one of the word senses and one of the terms;

analyze the semantic similarity values determined for respective ones of the word senses; and

select one of the word senses using the analysis, wherein the one of the word senses has an increased relevancy with respect to the terms of the documents of the cluster compared with the relevancies of others of the word senses.

34. The article of claim 33 wherein the programming is configured to cause the processing circuitry to determine, for individual ones of the word senses, a plurality of semantic similarity values with respect to individual ones of the terms.

35. The article of claim 33 wherein the programming is configured to cause the processing circuitry to select the one of the word senses having a greatest cumulative semantic similarity with respect to the terms.

36. The article of claim 33 wherein the programming is configured to cause the processing circuitry to, for individual ones of the terms, select the greatest semantic similarity values for individual ones of the word senses and, for individual ones of the word senses, to sum the selected greatest semantic similarity values for the respective word sense providing a cumulative value for the word sense to analyze the semantic similarity values, and to select the one of the word senses having the greatest cumulative value.

37. The article of claim 33 wherein the terms comprise minor terms and major terms, and the cluster label comprises a major term.

38. The article of claim 33 wherein the selected one of the word senses is indicative of contents of the documents of the subset which correspond to the cluster.

39. The article of claim 38 wherein the programming is configured to cause the processing circuitry to control a display to depict the cluster and the selected one of the word senses in a manner which is associated with the cluster.