Method and system for analyzing user-generated content

• US 7,930,302 B2
• Filed: 11/05/2007
• Issued: 04/19/2011
• Est. Priority Date: 11/22/2006
• Status: Active Grant

• Current Assignee: Intuit, Inc.
• Original Assignee: Intuit, Inc.
• Inventors: Bandaru, Nagaraju, Moyer, Eric D., Radhakrishna, Shrisha
• Primary Examiner(s): Stevens; Robert
• Application Number:
  US11/982,936
• Publication Number:
  US 20080133488A1
• Time in Patent Office: 1,261 Days
• Field of Search: 707/1, 707/3, 707/999.3, 707/737, 715/810
• US Class Current: 707/737
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06F 40/30 Semantic analysis

Abstract

A method and system for collecting and analyzing data found across multiple sites on the internet or stored in a self-contained or pre-loaded database, is disclosed which captures, extracts, analyzes, categorizes, synthesizes, summarizes and displays, in a customizable format, both the substance and sentiment embodied within user-generated content, such as comments or reviews, found across such sites and/or stored within such databases.

Claims

1. A method for analyzing user-generated content, comprising:

collecting, by a processor, a review of a business from a website;

extracting, by the processor, a sentence from the review;

identifying, processor and within the sentence, an attribute-value comprising a noun and a descriptive term describing the noun, wherein the descriptive term comprises an adjective and an adverb corresponding to the adjective;

obtaining, by the processor, an adjective score for the adjective from a rating system lookup table;

assigning, by the processor, a heuristic rule score to the adverb; and

calculating, by the processor, a sentence review score for the sentence by summing the adjective score and the heuristic rule score.

2. The method of claim 1, wherein the review is derived from a blog.

3. The method of claim 1, wherein the review is for a product sold by the business.

4. The method of claim 1, further comprising:

mapping the noun to a core attribute of the business;

identifying a plurality of sentence review scores including the sentence review score corresponding to the core attribute of the business; and

calculating a comprehensive review of the core attribute for the business based on the plurality of sentence review scores.

5. The method of claim 4, wherein the business is a restaurant, and wherein the core attribute is one selected from a group consisting of price, service, ambiance, and food.

6. The method of claim 1, further comprising:

collecting a plurality of reviews for the business;

calculating a frequency of occurrence of the attribute-value within the plurality of reviews; and

displaying the attribute-value according to the frequency of occurrence.

7. The method of claim 1, further comprising:

collecting a plurality of reviews for the business;

identifying a plurality of favorable attribute-values within the plurality of reviews, wherein the plurality of favorable attribute-values includes the attribute-value; and

displaying, on an advertisement, at least one of the plurality of favorable attribute-values.

8. The method of claim 1, further comprising:

receiving, from a user, a search query comprising the noun;

identifying, based on the search query, a plurality of sentence review scores comprising the sentence review score;

calculating, based on the plurality of sentence review scores and using a normalization algorithm, a plurality of normalized sentence review scores;

calculating a representative sentence score based on the plurality of normalized sentence review scores; and

outputting the representative sentence score to the user.

9. The method of claim 1, further comprising:

retrieving a plurality of attribute-values corresponding to the review and comprising the attribute-value; and

generating a summary of the review comprising a plurality of words from the plurality of attribute-values.

10. The method of claim 1, wherein the rating system lookup table uses a third party lexical reference system.

11. The method of claim 10, further comprising:

identifying a dictionary comprising a plurality of words; and

mapping the noun to a word of the plurality of words, wherein the noun is a synonym of the word, and wherein mapping the noun to the word is done by the third party lexical reference system.

12. A system for analyzing user-generated content, comprising:

a hardware processor;

a semantic crawler executing on the hardware processor and configured to collect a review of a business from a website;

a parser executing on the hardware processor and configured to extract a sentence from the review; and

a semantic processor executing on the hardware processor and configured to;

identify, within the sentence, an attribute-value comprising a noun and a descriptive term describing the noun, wherein the descriptive term comprises an adjective and an adverb corresponding to the adjective;

obtain an adjective score for the adjective from a rating system lookup table;

assign a heuristic rule score to the adverb; and

calculate a sentence review score for the sentence by summing the adjective score and the heuristic rule score.

13. The system of claim 12, wherein the business is a restaurant and wherein the semantic processor is further configured to:

categorize, based on the sentence review score, the attribute-value as one selected from a group consisting of positive, negative, and neutral.

14. The system of claim 12, wherein the semantic processor is further configured to:

collect a plurality of reviews for the business;

identify a plurality of favorable attribute-values from the plurality of reviews, wherein the plurality of favorable attribute-values includes the attribute-value; and

display, on an advertisement, at least one of the plurality of favorable attribute-values.

15. The system of claim 12, wherein the semantic processor is further configured to:

receive, from a user, a search query comprising the noun from a user;

identify, based on the search query, a plurality of sentence review scores comprising the sentence review score;

calculate, based on the plurality of sentence review scores and using a normalization algorithm, a plurality of normalized sentence review scores;

calculate a representative sentence score based on the plurality of normalized sentence review scores; and

output the representative sentence score to the user.

16. The system of claim 12, wherein the semantic processor is further configured to:

map the noun to a core attribute of the business;

identify a plurality of sentence review scores including the sentence review score corresponding to the core attribute of the business; and

calculate a comprehensive review of the core attribute of the business based on the plurality of sentence review scores.

17. The system of claim 12, wherein the semantic processor is further configured to:

collect a plurality of reviews for the business;

calculate a frequency of occurrence of the attribute-value within the plurality of reviews; and

display the attribute-value according to the frequency of occurrence.

18. A computer-readable storage medium storing a plurality of instructions for analyzing user-generated content, the plurality of instructions comprising functionality to:

collect a review of a business from a website;

extract a sentence from the review;

identify, within the sentence, an attribute-value comprising a noun and a descriptive term describing the noun, wherein the descriptive term comprises an adjective and an adverb corresponding to the adjective;

obtain an adjective score for the adjective from a rating system lookup table;

assign a heuristic rule score to the adverb; and

calculate a sentence review score for the sentence by summing the adjective score and the heuristic rule score.

19. The computer-readable storage medium of claim 18, wherein the plurality of instructions further comprise functionality to:

receive, from a user, a search query comprising the noun;

identify, based on the search query, a plurality of sentence review scores comprising the sentence review score;

calculate, based on the plurality of sentence review scores and using a normalization algorithm, a plurality of normalized sentence review scores;

calculate a representative sentence score based on the plurality of normalized sentence review scores; and

output the representative sentence score to the user.

20. The computer-readable storage medium of claim 18, wherein the plurality of instructions further comprise functionality to:

map the noun to a core attribute of the business;

identify a plurality of sentence review scores including the sentence review score corresponding to the core attribute of the business; and

calculate a comprehensive review of the core attribute of the business based on the plurality of sentence review scores.

21. The computer-readable storage medium of claim 18, wherein the plurality of instructions further comprise functionality to:

collect a plurality of reviews for the business;

calculate a frequency of occurrence of the attribute-value within the plurality of reviews; and

display the attribute-value according to the frequency of occurrence.