METHODS FOR IDENTIFYING DOCUMENTS RELATING TO A MARKET

• US 20090125381A1
• Filed: 11/05/2008
• Published: 05/14/2009
• Est. Priority Date: 11/07/2007
• Status: Abandoned Application

• Current Assignee: KPMG LLP (KPMG International Cooperative)
• Original Assignee: Wise Window, Inc.
• Inventors: Delepet, Rajiv
• Application Number:
  US12/265,107
• Publication Number:
  US 20090125381A1
• Time in Patent Office: Days
• US Class Current: 705/10
• CPC Class Codes: G06F 16/313 Selection or weighting of terms for indexing
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0201 Market modelling; Market analysis; Collecting market data
• CPC Class Codes: G06Q 30/0204 Market segmentation

Abstract

Methods of identifying web documents as relating to a market domain that would ordinarily be considered unrelated are presented. Market domain criteria can be defined that provide for classifying web documents as being related to the domain. The documents classified as related to the market domain form a training sample of documents used to establish correlations among brand term combinations found within the documents. If correlations are established among the terms in a combination, the term combinations can be assigned a similarity score indicating how similar the terms are considered to be. The term combinations can be used to search for additional web documents that could pertain to the market domain but would otherwise fail to satisfy the market domain criteria. The search results can be presented via a computer interface according to similarity scores.

Claims

1. A method of identifying web documents pertaining a market domain, the method comprising:

classifying a first group of web documents as relating to a specified market domain;

identifying a set of brand term combinations, where each combination includes a first brand term and a second, different brand term occurring within the first group of web documents;

establishing a term correlation between the first term and the second term for each combination based on usage of the first and the second terms within the first group of web documents;

assigning a similarity score to each combination as a function of the term correlation;

searching for a second group of web documents unrelated to the specified market domain using at least some of combinations; and

presenting the second group of web documents via a computer interface according the similarity scores of the brand term combinations.

2. The method of claim 1, wherein the step of classifying is performed by a human being.

3. The method of claim 1, wherein the step of classifying includes automatically identifying the first group of web documents as a function of a search term representing the specified market domain.

4. The method of claim 3, wherein the search term comprises data selected from the group consisting of image data and audio data.

5. The method of claim 3, further comprising automatically updating the first group of web documents.

6. The method of claim 1, wherein the first group of web documents are sourced from e-commerce sites.

7. The method of claim 1, wherein the first group of web documents comprises reviews.

8. The method of claim 1, wherein the second group of web documents comprises articles.

9. The method of claim 8, wherein the articles comprises at least one of the following types of articles:

a blog, a forum post, and a news story.

10. The method of claim 1, wherein the step of identifying the set of combinations includes ignoring terms on an exclusion list.

11. The method of claim 1, wherein the usage includes proximity of the first and the second term within individual ones of the first group of web documents.

12. The method of claim 1, wherein the usage includes frequency of occurrences of the first and the second terms within the first group of web documents.

13. The method of claim 1, wherein the usage includes a difference between digital data used to represent the first brand term and digital data used to represent the second brand term.

14. The method of claim 1, further comprising periodically updating the similarly score.

15. The method of claim 1, further comprising reducing the set of brand term combinations by eliminating combinations having similarity scores outside a threshold constraint.

16. The method of claim 1, wherein the step of presenting the second group of web documents includes providing a web services application program interface.

17. The method of claim 1, further comprising providing access to a history of the set of combinations having similarity scores for the specified market domain.