Topical sentiments in electronically stored communications

• US 8,041,669 B2
• Filed: 12/15/2010
• Issued: 10/18/2011
• Est. Priority Date: 09/30/2004
• Status: Expired due to Fees

• Current Assignee: Buzzmetrics, Ltd. (Nielsen Holdings plc)
• Original Assignee: Buzzmetrics, Ltd. (Nielsen Holdings plc)
• Inventors: Hurst, Matthew F., Nigam, Kamal P.
• Primary Examiner(s): Holmes; Michael B
• Application Number:
  US12/969,356
• Publication Number:
  US 20110093417A1
• Time in Patent Office: 307 Days
• Field of Search: 706/55
• US Class Current: 706/55
• CPC Class Codes: G06F 40/253 Grammatical analysis; Style critique
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising

Abstract

The present application presents methods for performing topical sentiment analysis on electronically stored communications employing fusion of polarity and topicality. The present application also provides methods for utilizing shallow NLP techniques to determine the polarity of an expression. The present application also provides a method for tuning a domain-specific polarity lexicon for use in the polarity determination. The present application also provides methods for computing a numeric metric of the aggregate opinion about some topic expressed in a set of expressions.

Claims

1. A system comprising:

a text mining tool to process text in an electronic document to identify a topical expression and a polar expression in the electronic document;

a text classifier to determine a topic of the topical expression and a polarity of the polar expression, to identify a relevance of the polar expression to the topical expression, and to generate a confidence score associated with the relevance of the polar expression to the topical expression; and

a user interface to display the topical expression with the polar expression and the confidence score to a user.

2. The system of claim 1, wherein the user interface is to display an aggregated plurality of topical expressions and corresponding polar expressions identified within the electronic document.

3. The system of claim 1, wherein the user interface is to display an aggregated plurality of topical expression and corresponding polar expressions identified from a plurality of electronic documents.

4. The system of claim 1, wherein the text mining tool is to perform a natural language processing analysis of the electronic document to identify a topical expression and a polar expression, and the text classifier is to perform a natural language processing analysis of the topical expression and the polar expression to identify a relevance of the polar expression to the topical expression and to generate a confidence score associated with the relevance of the polar expression to the topical expression.

5. The system of claim 1, wherein the text classifier comprises a statistical machine learning classifier.

6. The system of claim 5, wherein the statistical machine learning classifier is to execute a Winnow analysis with respect to the electronic document.

7. The system of claim 1, wherein the text classifier comprises a rules-based classifier and is to analyze the polar expression using a shallow natural language processing technique.

8. A method comprising:

processing text in electronic content to identify a topical expression and a polar expression in the electronic content;

determining a topic of the topical expression and a polarity of the polar expression;

identifying a relevance of the polar expression to the topical expression;

generating a confidence score associated with the relevance of the polar expression to the topical expression; and

displaying the topical expression with the polar expression and the confidence score to a user.

9. The method of claim 8, further comprising displaying an aggregated plurality of topical expressions and corresponding polar expressions identified within the electronic content.

10. The method of claim 8, further comprising displaying an aggregated plurality of topical expression and corresponding polar expressions identified from a plurality of electronic content.

11. The method of claim 8, wherein processing text further comprises performing a natural language processing analysis of the electronic content to identify the topical expression and the polar expression, and wherein identifying further comprises performing a natural language processing analysis of the topical expression and the polar expression to identify a relevance of the polar expression to the topical expression.

12. The method of claim 8, wherein identifying further comprises applying a statistical machine learning classification to identify a relevance of the polar expression to the topical expression.

13. The method of claim 8, wherein the identifying further comprises applying a rules-based classification with a shallow natural language processing technique to analyze the polar expression.

14. A tangible computer readable storage medium including executable program instructions which, when executed by a computer processor, cause the computer to implement a system comprising:

a text mining tool to process text in an electronic document to identify a topical expression and a polar expression in the electronic document;

a text classifier to determine a topic of the topical expression and a polarity of the polar expression, to identify a relevance of the polar expression to the topical expression; and

a user interface to display the topical expression with the polar expression to a user.

15. The computer readable storage medium of claim 14, wherein the text classifier is to generate a confidence score associated with the relevance of the polar expression to the topical expression.

16. The computer readable storage medium of claim 14, wherein the user interface is to display an aggregated plurality of topical expression and corresponding polar expressions identified from a plurality of electronic documents.

17. The computer readable storage medium of claim 14, wherein the text mining tool is to perform a natural language processing analysis of the electronic document to identify a topical expression and a polar expression and the text classifier is to perform a natural language processing analysis of the topical expression and the polar expression to identify a relevance of the polar expression to the topical expression, and to generate a confidence score associated with the relevance of the polar expression to the topical expression.

18. The computer readable storage medium of claim 14, wherein the text classifier comprises a rules-based classifier to apply a statistical generator model to analyze the polar expression in relation to the topical expression.