Systems and methods for creating an interest profile for a user

• US 9,116,979 B2
• Filed: 12/02/2011
• Issued: 08/25/2015
• Est. Priority Date: 06/22/2011
• Status: Active Grant

• Current Assignee: Rogers Communications Inc.
• Original Assignee: Rogers Communications Inc.
• Inventors: Miao, Yingbo, Xu, Liqin, Lee, Hyun Chul
• Primary Examiner(s): Dang, Thanh-Ha
• Application Number:
  US13/309,661
• Publication Number:
  US 20120330975A1
• Time in Patent Office: 1,362 Days
• Field of Search: 707/749, 707/732, 707/733, 707/734
• US Class Current: 1/1
• CPC Class Codes: G06F 16/285 Clustering or classification
• CPC Class Codes: G06F 16/337 Profile generation, learning or modification
• CPC Class Codes: G06F 16/435 Filtering based on additional data, e.g. user or group profiles
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques

Abstract

Profiling systems and methods of creating and using user interest profiles are described. In some example embodiments, the method includes: creating a topic set which includes topics which are organized in a hierarchical structure which includes a plurality of topic levels including an upper topic level and a lower topic level, each topic in the lower topic level being a subtopic of at least one of the topics in the upper topic level; monitoring interest in a plurality of documents for a user to identify one or more documents-of-interest to the user; and based on the monitored interest for the user, creating an interest profile for the user by determining a measure of topical interest for the user for at least one of the topics at the upper topic level and for a subtopic of that topic, the subtopic being at the lower topic level.

Claims

1. A method implemented by a processor of a profiling system, the method comprising:

based on a document corpus containing a plurality of documents, creating a topic set, at least some of the topics in the topic set being organized in a hierarchical structure which includes a plurality of topic levels including an upper topic level and a lower topic level, each topic in the lower topic level being a subtopic of at least one of the topics in the upper topic level;

monitoring interest in a plurality of documents for a user to identify one or more documents-of-interest to the user; and

based on the monitored interest for the user, creating an interest profile for the user by;

determining a measure of topical interest for the user for at least one of the topics at the upper topic level and for a subtopic of that topic, the subtopic being at the lower topic level; and

determining, by the processor, a per-topic heuristic feature profile for the user, the per-topic heuristic feature profile indicating non-content related features which affect a user'"'"'s interest in a document by specifying, for at least one topic of the topic set, an effect of document type on the user'"'"'s interest in documents associated with the at least one topic, wherein the document type is one or more of;

a blog post, a micro-blog post, a comment, or a news article.

2. The method of claim 1, further comprising:

determining a measure of predicted interest in a document based on the interest profile for the user.

3. The method of claim 1, wherein each topic in the topic set is represented by a bag-of-words vector.

4. The method of claim 3, wherein the bag-of words vector is generated based on the content of documents related to the topic.

5. The method of claim 1, wherein determining a measure of topical interest comprises:

comparing the content of each document-of-interest to one of the topics.

6. The method of claim 5, wherein comparing the content of each document-of-interest to one of the topics comprises:

obtaining a measure of similarity between each document-of-interest and the topic.

7. The method of claim 6, wherein the measure of topical interest for the user and one of the topics is determined based on the measure of similarity between each document-of-interest and the topic.

8. The method of claim 6, wherein the measure of similarity is a cosine similarity between the document-of-interest and the topic.

9. The method of claim 1, further comprising:

determining a measure of significance for the user for at least one of the topics at the upper topic level and for a subtopic of that topic, the subtopic being at the lower topic level.

10. The method of claim 9, wherein the measure of significance is determined for each topic in the topic set.

11. The method of claim 9, wherein the measure of significance indicates the relative importance of the topic at the upper topic level and the subtopic of that topic for the user.

12. The method of claim 11, wherein the measure of significance is determined by applying a predicted interest formula, which finds a measure of predicted interest, to documents-of-interest and by setting the measure of predicted interest for documents-of-interest to a maximum value, the predicted interest formula including the measure of significance.

13. The method of claim 1, wherein the heuristic feature profile specifies per-topic interest in any one or more of the following document types:

blog posts, comments, news articles, or micro-blog posts.

14. The method of claim 1, wherein monitoring interest in a plurality of documents for a user comprises:

receiving direct user-generated input from the user indicating whether that user was interested in a document.

15. A profiling system comprising:

a processor; and

a memory coupled to the processor, the memory storing processor executable instructions which, when executed by the processor cause the processor to;

based on a document corpus containing a plurality of documents, create a topic set, at least some of the topics in the topic set being organized in a hierarchical structure which includes a plurality of topic levels including an upper topic level and a lower topic level, each topic in the lower topic level being a subtopic of at least one of the topics in the upper topic level;

monitor interest in a plurality of documents for a user to identify one or more documents-of-interest to the user; and

based on the monitored interest for the user, create an interest profile for the user by;

determining a measure of topical interest for the user for at least one of the topics at the upper topic level and for a subtopic of that topic, the subtopic being at the lower topic level; and

determining a per-topic heuristic feature profile for the user, the per-topic heuristic feature profile indicating non-content related features which affect a user'"'"'s interest in a document by specifying, for at least one topic of the topic set, an effect of document type on the user'"'"'s interest in documents associated with the at least one topic, wherein the document type is one or more of;

a blog post, a micro-blog post, a comment, or a news article.

16. The profiling system of claim 15, wherein the processor executable instructions are further configured to cause the processor to:

determine a measure of predicted interest in a document based on the interest profile for the user.

17. The profiling system of claim 16, wherein determining a measure of topical interest comprises:

comparing the content of each document-of-interest to one of the topics.

18. The profiling system of claim 17, wherein comparing the content of each document-of-interest to one of the topics comprises:

obtaining a measure of similarity between each document-of-interest and the topic.

19. The profiling system of claim 18, wherein the measure of topical interest for the user and one of the topics is determined based on the measure of similarity between each document-of-interest and the topic.

20. The profiling system of claim 18, wherein the measure of similarity is a cosine similarity between the document-of-interest and the topic.

21. The profiling system of claim 15, wherein each topic in the topic set is represented by a bag-of-words vector.

22. The profiling system of claim 21, wherein the bag-of words vector is generated based on the content of documents related to the topic.

23. The profiling system of claim 15, wherein the processor executable instructions are further configured to cause the processor to:

determine a measure of significance for the user for at least one of the topics at the upper topic level and for a subtopic of that topic, the subtopic being at the lower topic level.