Methods and apparatus to classify text communications

• US 8,909,645 B2
• Filed: 01/10/2013
• Issued: 12/09/2014
• Est. Priority Date: 08/14/2009
• Status: Active Grant

• Current Assignee: Buzzmetrics, Ltd. (Nielsen Holdings plc)
• Original Assignee: Buzzmetrics, Ltd. (Nielsen Holdings plc)
• Inventors: Eden, Tal, Greitzer, Eliyahu, Krichman, Yakir, Fuks, Michael
• Primary Examiner(s): Jami, Hares
• Application Number:
  US13/708,803
• Publication Number:
  US 20130138430A1
• Time in Patent Office: 698 Days
• Field of Search: 707/705, 707/722, 707/736, 707/737, 707/740, 707/753, 707/999.1, 707/999.7
• US Class Current: 707/737
• CPC Class Codes: G06F 16/355 Class or cluster creation or modification
• CPC Class Codes: G06F 40/40 Processing or translation of natural language natural language analysis G06F40/20; semantic analysis G06F40/30

Abstract

Methods and apparatus to classify text communications are disclosed. An example method includes determining a first score indicating a likelihood that a text belongs to a first classification mode by combining a first sentence score and a second sentence score retrieved from an index, the first sentence score indicating a probability that a first sentence in the text belongs to the first classification mode, the second sentence score indicating that a second sentence following the first sentence belongs to the first classification mode, determining a second score indicating a likelihood that the text belongs to a second classification mode, comparing the first score to the second score, classifying the text as the first classification mode when the first score is greater than the second score, and determining a confidence level that the text belongs to the first classification mode by dividing the first score by the second score.

Claims

1. A method to classify text communications, the method comprising:

determining a first score indicating a likelihood that a text belongs to a first classification mode by combining a first sentence score and a second sentence score retrieved from an index, the first sentence score indicating a first probability that a first sentence in the text belongs to the first classification mode, the second sentence score indicating a second probability that a second sentence following the first sentence belongs to the first classification mode;

determining a second score indicating a likelihood that the text belongs to a second classification mode;

comparing the first score to the second score;

classifying the text as the first classification mode when the first score is greater than the second score; and

determining a confidence level that the text belongs to the first classification mode by dividing the first score by the second score.

2. The method as defined in claim 1, wherein the first score is determined by:

determining a first conditional probability of a first feature occurring in the text given that the text belongs to the first classification mode, wherein the first feature is included in the text;

determining a second conditional probability of a second feature occurring in the text given that the text belongs to the first classification mode, wherein the second feature is included in the text;

determining a third probability of the first classification mode occurring; and

determining a product by multiplying the first conditional probability, the second conditional probability and the third probability of the first classification mode to determine the first score.

3. The method as defined in claim 2, wherein the first score is determined by determining a logarithm of the product.

4. The method as defined in claim 3, wherein multiplying and determining the logarithm of the product is performed by adding a logarithm of the first conditional probability to (1) a logarithm of the second conditional probability and (2) the logarithm of the third probability of the classification mode.

5. The method as defined in claim 2, wherein the first conditional probability is determined by dividing a number of times the first feature appeared in training text that was classified as the first classification mode by a total number of times features appeared in the training text that was classified as the first classification mode.

6. The method as defined in claim 2, wherein the third probability of the classification mode is determined by dividing the number of documents in the text that belonged to the first classification mode by the number of documents in the text.

7. The method as defined in claim 1, wherein the first classification mode is a sentiment classification.

8. The method as defined in claim 1, wherein the first classification mode comprises at least one of a positive classification, a negative classification, a mixed classification or a no opinion classification.

9. A tangible computer readable storage device or storage disc comprising instructions that, when executed, cause a machine to at least:

determine a first score indicating a likelihood that a text belongs to a first classification mode by combining a first sentence score and a second sentence score retrieved from an index, the first sentence score indicating a first probability that a first sentence in the text belongs to the first classification mode, the second sentence score indicating a second probability that a second sentence following the first sentence belongs to the first classification mode;

determine a second score indicating a likelihood that the text belongs to a second classification mode;

compare the first score to the second score;

classify the text is one of the first classification mode or the second classification mode based on the comparison of the first score and the second score; and

determine a confidence level that the text belongs to the first classification mode by dividing the first score by the second score.

10. The tangible computer readable storage device or storage disc as defined in claim 9, wherein the instructions, when executed, determine the first score by:

determining a first conditional probability of a first feature occurring in the text given that the text belongs to a first classification mode, wherein the first feature is included in the text;

determining a second conditional probability of a second feature occurring in the text given that the text belongs to the first classification mode, wherein the second feature is included in the text;

determining a third probability of the first classification mode occurring; and

determining a product by multiplying the first conditional probability, the second conditional probability and the third probability of the first classification mode to determine the first score.

11. The tangible computer readable storage device or storage disc as defined in claim 10, wherein the instructions, when executed, determine the first score by determining a logarithm of the product.

12. The tangible computer readable storage device or storage disc as defined in claim 11, wherein the instructions, when executed, determine the logarithm of the product by adding a logarithm of the first conditional probability to (1) a logarithm of the second conditional probability and (2) the logarithm of the third probability of the classification mode.

13. The tangible computer readable storage device or storage disc as defined in claim 11, wherein the instructions, when executed, determine the first conditional probability by dividing a number of times the first feature appeared in training text that was classified as the first classification mode by a total number of times features appeared in the training text that was classified as the first classification mode.

14. The tangible computer readable storage device or storage disc as defined in claim 10, wherein the instructions, when executed, determine the third probability of the first classification mode occurring by determining the number of documents in the text that belonged to the first classification mode and dividing by the number of documents in the text.

15. The tangible computer readable storage device or storage disc as defined in claim 9, wherein the first classification mode is a sentiment classification.

16. The tangible computer readable storage device or storage disc as defined in claim 9, wherein the first classification mode comprises at least one of a positive classification, a negative classification, a mixed classification or a no opinion classification.

17. An apparatus to classify text communications, the apparatus comprising:

a statistics generator to determine a first score indicating a first likelihood that a text belongs to a first classification mode by combining a first sentence score and a second sentence score retrieved from an index, the first sentence score indicating a first probability that a first sentence in the text belongs to the first classification mode, the second sentence score indicating a second probability that a second sentence following the first sentence belongs to the first classification mode, and the statistics generator is to determine a second score indicating a second likelihood that the text belongs to a second classification mode; and

a search engine to compare the first score to the second score, classify the text as the first classification mode when the first score is greater than the second score, and determine a confidence level that the text belongs to the first classification mode by dividing the first score by the second score, at least one of the statistics generator or the search engine being implemented via a processor.

18. The apparatus as defined in claim 17, wherein the statistics generator is to determine the first score by:

determining a first conditional probability of a first feature occurring in the text given that the text belongs to a first classification mode, wherein the first feature is included in the text;

determining a second conditional probability of a second feature occurring in the text given that the text belongs to the first classification mode, wherein the second feature is included in the text;

determining a third probability of the first classification mode occurring; and

determining a product by multiplying the first conditional probability, the second conditional probability and the third probability of the first classification mode to determine the first score.

19. The apparatus as defined in claim 18, wherein the statistics generator is to determine the first score by determining a logarithm of the product.

20. The apparatus as defined in claim 19, wherein the statistics generator is to determine the logarithm of the product by adding a logarithm of the first conditional probability to (1) a logarithm of the second conditional probability and (2) the logarithm of the third probability of the first classification mode.

21. The apparatus as defined in claim 18, wherein the statistics generator is to determine the first conditional probability by dividing a number of times the first feature appeared in training text that was classified as the first classification mode by a total number of times features appeared in the training text that was classified as the first classification mode.

22. The apparatus as defined in claim 18, wherein the statistics generator is to determine the third probability of the first classification mode by determining the number of documents in the text that belonged to the first classification mode and dividing by the number of documents in the text.

23. The apparatus as defined in claim 17, wherein the first classification mode is a sentiment classification.

24. The apparatus as defined in claim 17, wherein the first classification mode comprises at least one of a positive classification, a negative classification, a mixed classification or a no opinion classification.