Method and apparatus for performing context-based entity association

• US 8,639,688 B2
• Filed: 11/12/2009
• Issued: 01/28/2014
• Est. Priority Date: 11/12/2009
• Status: Expired due to Fees

• Current Assignee: Palo Alto Research Center, Inc. (Xerox Holdings Corp.), Xerox Corporation (Xerox Holdings Corp.)
• Original Assignee: Palo Alto Research Center, Inc. (Xerox Holdings Corp.)
• Inventors: Staddon, Jessica N., Chow, Richard, Golle, Philippe J. P., Purvis, Lisa S.
• Primary Examiner(s): Lee, Wilson
• Assistant Examiner(s): Le, Jessica N
• Application Number:
  US12/617,370
• Publication Number:
  US 20110113028A1
• Time in Patent Office: 1,538 Days
• Field of Search: None
• US Class Current: 707/723
• CPC Class Codes: G06Q 10/00 Administration; Management
• CPC Class Codes: G06Q 30/0241 Advertisements

Abstract

A targeted advertising system performs context-based association mining using a publicly available corpus to identify a product or brand name that, under a given context, is associated with a product or brand being marketed. The system analyzes documents within the publicly available corpus that are associated with the given context, and identifies products or brand names that have a high association to the product or brand being marketed. The system can also analyze the publicly available corpus to determine contextual information which is correlated to two or more products or brand names. This contextual information includes a set of terms that facilitates filtering the publicly available corpus into an optimal set of documents that has a high association to a desired market category or demographic.

Claims

1. A computer-executed method that identifies, for an initial product to advertise, a product that is likely to be of interest to consumers of the initial product, the method comprising:

receiving a first product identifier, which includes a set of terms associated with the initial product;

determining optimized contextual information for a target market category of the initial product, wherein the optimized contextual information includes a set of terms associated with the initial product or the target market category;

performing association mining, by searching a set of documents in a publicly available corpus that mention the first product identifier and the optimized contextual information, to determine a target product that is likely to be of interest to consumers of the initial product; and

in response to determining the target product by performing association mining on the set of documents, determining a confidence value for the target product using documents related to the optimized contextual information, wherein computing the confidence value involves computing a ratio of the set of documents that mention the initial product, the target product, and the optimized contextual information, to the set of documents that mention the initial product and the optimized contextual information.

2. The computer-executed method of claim 1, wherein performing association mining comprises:

selecting the second product identifier from a set of product identifiers associated with the target market category, wherein the second product identifier includes a first set of terms associated with a related product in the target market category that is likely to be of interest to consumers of the initial product;

generating a second set of terms for the initial product, which includes one or more terms selected from at least the first product identifier, and includes one or more terms selected from at least the contextual information; and

determining the confidence value for the second product identifier by computing, from a set of documents in the publicly available corpus that include the second set of terms, the percentage of documents that also include the second product identifier.

3. The computer-executed method of claim 2, wherein performing association mining further comprises ranking the set of product identifiers associated with the target market category based in part on a corresponding set of confidence values.

4. The computer-executed method of claim 2, wherein the publicly available corpus includes the Internet;

and

wherein determining the confidence value comprises;

performing an Internet search query based in part on the second set of terms to determine a first hit count, wherein the first hit count indicates a number of Internet web pages that mention the initial product and optionally mention the target market category;

performing an Internet search query based in part on the second set of terms, the second product identifier, and the contextual information to determine a second hit count, wherein the second hit count indicates a number of Internet web pages that mention the initial product, the determined product, and optionally mention the target market category; and

computing the ratio of the second hit count to the first hit count.

5. The computer-executed method of claim 2, wherein generating the second set of terms which includes the one or more terms selected from at least the contextual information comprises:

identifying one or more terms associated with the target market category which are not included in the contextual information; and

inserting the identified terms into the second set of terms.

6. The computer-executed method of claim 2, wherein generating the second set of terms which includes the one or more terms selected from at least the contextual information comprises:

identifying one or more terms in the contextual information which have a weak association with the target market category; and

generating the second set of terms without including the identified terms.

7. The computer-executed method of claim 2, wherein producing the first set of terms associated with the related product comprises generating a third set of terms, which includes one or more terms selected from the second product identifier, and includes zero or more terms selected from at least the contextual information.

8. The computer-executed method of claim 2, further comprising:

generating a plurality of term combinations that include contextual information;

performing an Internet search query for each term combination to determine a hit count for each term combination;

determining a group of term combinations that have substantially similar hit counts;

determining an optimized term-count for the contextual information based on the group of term combinations; and

selecting a set of terms for the optimized contextual information based on the optimized term-count.

9. The computer-executed method of claim 1, wherein determining contextual information comprises:

receiving a first set of terms associated with the target market category;

generating a second set of terms which includes a set of synonyms and hypernyms for one or more terms in the first set of terms; and

producing a third set of terms associated with the contextual information which includes the first set of terms and the second set of terms.

10. The computer-executed method of claim 1, further comprising receiving a second product identifier, which includes a set of terms associated with the determined product;

wherein determining contextual information comprises performing association mining to determine a third set of terms, which includes a set of keywords that are associated with the initial product and are associated with the determined product, by;

performing a first search query through the publicly available corpus to determine a first set of terms which are associated with the initial product;

performing a second search query through the publicly available corpus to determine a second set of terms which are associated with the determined product; and

determining the third set of terms which occur in the first set of terms and occur in the second set of terms.

12. The computer-readable storage medium of claim 10, wherein performing association mining comprises:

selecting the second product identifier from a set of product identifiers associated with the target market category, wherein the second product identifier includes a first set of terms associated with a related product in the target market category that is likely to be of interest to consumers of the initial product;

generating a second set of terms for the initial product, which includes one or more terms selected from at least the first product identifier, and includes one or more terms selected from at least the contextual information; and

determining the confidence value for the second product identifier by computing, from a set of documents in the publicly available corpus that include the second set of terms, the percentage of documents that also include the second product identifier.

13. The computer-readable storage medium of claim 12, wherein performing association mining further comprises ranking the set of product identifiers associated with the target market category based in part on a corresponding set of confidence values.

14. The computer-readable storage medium of claim 12, wherein the publicly available corpus includes the Internet;

and

wherein determining the confidence value comprises;

performing an Internet search query based in part on the second set of terms to determine a first hit count, wherein the first hit count indicates a number of Internet web pages that mention the initial product and optionally mention the target market category;

performing an Internet search query based in part on the second set of terms, the second product identifier, and the contextual information to determine a second hit count, wherein the second hit count indicates a number of Internet web pages that mention the initial product, the determined product, and optionally mention the target market category; and

computing the ratio of the second hit count to the first hit count.

15. The computer-readable storage medium of claim 12, wherein generating the second set of terms which includes the one or more terms selected from at least the contextual information comprises:

identifying one or more terms associated with the target market category which are not included in the contextual information; and

inserting the identified terms into the second set of terms.

16. The computer-readable storage medium of claim 12, wherein generating the second set of terms which includes the one or more terms selected from at least the contextual information comprises:

identifying one or more terms in the contextual information which have a weak association with the target market category; and

generating the second set of terms without including the identified terms.

17. The computer-readable storage medium of claim 12, wherein producing the first set of terms associated with the related product comprises generating a third set of terms, which includes one or more terms selected from the second product identifier, and includes zero or more terms selected from at least the contextual information.

11. A computer-readable storage medium storing instructions that when executed by a computer cause the computer to perform a method, the method comprising:

receiving a first product identifier, which includes a set of terms associated with an initial product to advertise;

determining optimized contextual information for a target market category of the initial product, wherein the optimized contextual information includes a set of terms associated with the initial product or the target market category;

performing association mining, by searching a set of documents in a publicly available corpus that mention the first product identifier and the optimized contextual information, to determine a target product that is likely to be of interest to consumers of the initial product; and

in response to determining the target product by performing association mining on the set of documents, determining a confidence value for the target product using documents related to the optimized contextual information, wherein computing the confidence value involves computing a percentage ratio of the set of documents that mention the initial product, the target product, and the optimized contextual information, to the set of documents that mention the initial product and the optimized contextual information.

18. The computer-readable storage medium of claim 11, wherein determining contextual information comprises:

receiving a first set of terms associated with the target market category;

generating a second set of terms which includes a set of synonyms and hypernyms for one or more terms in the first set of terms; and

producing a third set of terms associated with the contextual information which includes the first set of terms and the second set of terms.

19. The computer-readable storage medium of claim 11, further comprising receiving a second product identifier, which includes a set of terms associated with the determined product;

wherein determining contextual information comprises performing association mining to determine a set of keywords, which are associated with the initial product and are associated with the determined product, by;

performing a first search query through the publicly available corpus to determine a first set of terms which are associated with the initial product;

performing a second search query through the publicly available corpus to determine a second set of terms which are associated with the determined product; and

determining a third set of terms which occur in the first set of terms and occur in the second set of terms.

20. An apparatus for identifying, for an initial product to advertise, a product that is likely to be of interest to consumers of the initial product, comprising:

a processor;

a memory;

a receiving mechanism configured to receive a first product identifier, which includes a set of terms associated with the initial product; and

an association mining mechanism configured to;

determine optimized contextual information for a target market category of the initial product, wherein the optimized contextual information includes a set of terms associated with the initial product or the target market category;

perform association mining, by searching a set of documents in a publicly available corpus that mention the first product identifier and the optimized contextual information, to determine a target product that is likely to be of interest to consumers of the initial product; and

in response to determining the target product by performing association mining on the set of documents, determine a confidence value for the target product using documents related to the optimized contextual information, wherein computing the confidence value involves computing a ratio of the set of documents that mention the initial product, the target product, and the optimized contextual information, to the set of documents that mention the initial product and the optimized contextual information.

21. The apparatus of claim 20, wherein the association mining mechanism is further configured to perform association mining by:

selecting the second product identifier from a set of product identifiers associated with the target market category, wherein the second product identifier includes a first set of terms associated with a related product in the target market category that is likely to be of interest to consumers of the initial product;

generating a second set of terms for the initial product, which includes one or more terms selected from at least the first product identifier, and includes one or more terms selected from at least the contextual information; and

determining the confidence value for the second product identifier by computing, from a set of documents in the publicly available corpus that include the second set of terms, the percentage of documents that also include the second product identifier.