Method and system for automating the analysis of word frequencies

• US 7,131,117 B2
• Filed: 09/04/2002
• Issued: 10/31/2006
• Est. Priority Date: 09/04/2002
• Status: Expired due to Term

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: SBC Properties, L.P. (AT&T, Inc.)
• Inventors: Mills, Scott H., Joseph, Kurt M., Bushey, Robert R., Martin, John M., Knott, Benjamin A., Pasquale, Theodore B.
• Primary Examiner(s): Dam, Tuan
• Assistant Examiner(s): TECKLU, ISAAC TUKU
• Application Number:
  US10/234,398
• Publication Number:
  US 20040044950A1
• Time in Patent Office: 1,518 Days
• Field of Search: 717/141, 707/3, 707/4, 707/102, 715/531, 704/10
• US Class Current: 717/141
• CPC Class Codes: G06F 40/247 Thesauruses; Synonyms

Abstract

A method and system for automating the analysis of word frequencies includes a frequency system automatically analyzing a plurality of statements, a count engine, and a cluster engine. The count engine allows for the counting of unique words in the statements and the determination of a frequency of occurrence for each unique word. The frequency system further includes a phrase file allowing for the count engine to specify groups of words as single unique words and a synonym file allowing for the count engine to group one or more words together in synonym groups to be specified as single unique words. The cluster engine locates a plurality of clusters in the statements and determines a cluster frequency of occurrence for each of the clusters. The automated analysis of the statements allows for cost savings, more efficient use of time, and more reliable and consistent word frequency results.

Claims

1. A method for an automated analysis of word frequencies, the method comprising:

collecting a plurality of opening statements including a plurality of words, the plurality of words including multiple different transcriptions of a particular word or group of words;

specifying a synonym group including the multiple different transcriptions of the particular word or group of words;

establishing a cluster size;

removing a plurality of punctuation from the opening statements;

automatically counting each unique word in the opening statements, wherein the multiple different transcriptions of the particular word or group of words specified in the synonym group are treated as a single unique word;

determining a frequency of occurrence for each of the unique words in the opening statements;

automatically locating a plurality of clusters in the opening statements;

determining a cluster frequency of occurrence for each of the clusters;

storing each of the unique words and a corresponding frequency of occurrence and each of the clusters and a corresponding cluster frequency of occurrence; and

creating an output file sorted by the frequency of occurrence and the cluster frequency of occurrence and including each unique word, the corresponding frequency of occurrence, each cluster, and the corresponding cluster frequency of occurrence.

2. A method for automating an analysis of word frequencies, the method comprising:

accessing a plurality of transcribed statements including a plurality of words, the plurality of words including multiple different transcriptions of a particular word or group of words;

specifying a synonym group including the multiple different transcriptions of the particular word or group of words;

automatically analyzing the words in the statements;

determining a frequency of occurrence for each unique word in the statements, wherein the multiple different transcriptions of the particular word or group of words specified in the synonym group are treated as a single unique word; and

creating an output file including each of the unique words and a corresponding frequency of occurrence.

3. The method of claim 2 wherein the statements comprise customer opening statements.

4. The method of claim 2 wherein the output file comprises a frequency table.

5. The method of claim 2 wherein automatically analyzing the words in the statements comprises removing a plurality of punctuation from the statements.

6. The method of claim 5 wherein removing the punctuation from the statements comprises leaving the punctuation within each of the words.

7. The method of claim 2 wherein automatically analyzing the words in the statements comprises counting each unique word in the statements.

8. The method of claim 7 wherein counting each unique word in the statements comprises counting each of the groups of words as a single unique word.

9. The method of claim 7 wherein counting each of the unique words in the statements comprises counting the multiple different transcriptions of the particular word or group of words specified in the synonym group as a single unique word.

10. The method of claim 2 wherein automatically analyzing the words in the statements comprises:

establishing a cluster size; and

automatically locating a plurality of clusters in the statements.

11. The method of claim 10 further comprising determining a cluster frequency of occurrence for each of the clusters.

12. The method of claim 10 further comprising the output file sorted by cluster frequency of occurrence and including each of the clusters and a corresponding cluster frequency of occurrence.

13. The method of claim 2 further comprising utilizing one or more of the unique words having a high frequency occurrence in one or more customer-centric menu prompts.

14. A system for automating an analysis of word frequencies, the system comprising:

a plurality of statements including a plurality of words, the plurality of words including multiple different transcriptions of a particular word or group of words;

a synonym file specifying a synonym group including the multiple different transcriptions of the particular word or group of words;

a count engine operable to analyze the statements and determine a frequency of occurrence for each unique word in the statements, wherein the multiple different transcriptions of the particular word or group of words specified in the synonym group are treated as a single unique word; and

a cluster engine associated with the count engine, the cluster engine operable to locate a plurality of clusters in the statements and determine a cluster frequency of occurrence for each of the clusters.

15. The system of claim 14 wherein the statements comprise a plurality of opening statements.

16. The system of claim 14 wherein the count engine is further operable to remove a plurality of punctuation from the statements.

17. The system of claim 16 wherein the count engine removing the punctuation from the statements comprises leaving the punctuation within each of the words.

18. The system of claim 14 wherein the count engine is further operable to generate an output file including each of the unique words and a corresponding frequency of occurrence.

19. The system of claim 14 wherein the count engine counts each of the unique words in the statements.

20. The system of claim 14 further comprising:

a phrase file including one or more groups of words with each group specified as a unique word.

21. The system of claim 20 wherein the count engine counts each group of words in the phrase file as a single unique word and each word in the synonym group as a single unique word.

22. The system of claim 14 wherein a cluster comprises a grouping of two or more unique words occurring together in the statements.

23. The system of claim 14 wherein the cluster engine is operable to generate an output file including each of the clusters and a corresponding cluster frequency of occurrence.

24. The system of claim 14 further comprising a graphical user interface associated with the count engine and the cluster engine, the graphical user interface including a cluster control and operable to allow for selection of a phrase file and a synonym file.

25. A Computer program product for automating an analysis of word frequencies, the computer program product;

and embodied in a computer-readable storage medium and operable to;

access a plurality of transcribed statements including a plurality of words, the plurality of words including multiple different transcriptions of a particular word or group of words;

specifying a synonym group including the multiple different transcriptions of the particular word or group of words;

automatically analyze the words in the statements;

determine a frequency of occurrence for each unique word in the statements, wherein the multiple different transcriptions of the particular word or group of words specified in the synonym group are treated as a single unique word; and

create an output file including each of the unique words and a corresponding frequency of occurrence.

26. The medium of claim 25 wherein the output file comprises a frequency table.

27. The medium of claim 25 wherein automatically analyzing the words in the statements comprises removing a plurality of punctuation from the statements.

28. The medium of claim 25 wherein automatically analyzing the words in the statements comprises counting each unique word in the statements.

29. The medium of claim 28 wherein counting each of the unique words in the statements comprises counting a group of words as a single unique word.

30. The medium of claim 28 wherein counting each of the unique words in the statements comprises counting the words in the synonym group as a single unique word.

31. The medium of claim 25 wherein automatically analyzing the words in the statements comprises:

establishing a cluster size; and

automatically locating a plurality of clusters in the statements.

32. The medium of claim 31 further comprising determining a cluster frequency of occurrence for each of the clusters.

33. The medium of claim 25 further operable to store the unique words and the corresponding frequency of occurrence in a database.