Providing context for web articles

• US 8,630,972 B2
• Filed: 06/21/2008
• Issued: 01/14/2014
• Est. Priority Date: 06/21/2008
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Gamon, Michael, Basu, Sumit, Belenko, Dmitriy A., Konig, Arnd C., Hurst, Matthew F., Fisher, Danyel A
• Primary Examiner(s): Chaki, Kakali
• Assistant Examiner(s): COUGHLAN, PETER D
• Application Number:
  US12/143,765
• Publication Number:
  US 20090319449A1
• Time in Patent Office: 2,033 Days
• Field of Search: 706/58
• US Class Current: 706/58
• CPC Class Codes: G06F 16/94 Hypermedia Hyperlinking G06F40/134

Abstract

An overwhelming number of articles are available everyday via the internet. Unfortunately, it is impossible to peruse more than a handful, and it is difficult to ascertain an article'"'"'s social context. The techniques disclosed herein address this problem by harnessing implicit and explicit contextual information from social media. By extracting text surrounding a hyperlink to an article in a post and assessing the article as a function of content surrounding the hyperlink, an article'"'"'s social context is determined and presented. Additionally, articles that are sufficiently similar in content may be grouped to establish a many-to-one relationship between posts and an article, creating a more accurate assessment.

Claims

1. A method for assessing an emotional sentiment related to a topic, comprising:

identifying first social media content comprising a first link to a first article associated with the topic;

assessing an emotional sentiment related to the first article as a function of one or more terms in the first social media content;

identifying second social media content comprising a second link to a second article associated with the topic and comprising content similar to the first article, the second article comprising content similar to the first article when S(a_i,a_j)/size(a_i) exceeds a predetermined threshold where S(a_i,a_j) is a number of terms of a largest set of infrequent terms between the first article (a_i) and the second article (a_j) and size (a_i) is the number of terms in the larger of articles a_i and a_j;

assessing an emotional sentiment related to the second article as a function of one or more terms in the second social media content; and

aggregating the emotional sentiment related to the first article with the emotional sentiment related to the second article to assess the emotional sentiment related to the topic,

at least some of the method implemented at least in part via a processing unit.

2. The method of claim 1, comprising providing for display in a user interface at least one of:

a first graphical representation of the emotional sentiment related to the first article;

a second graphical representation of the emotional sentiment related to the second article;

or

a third graphical representation of the emotional sentiment related to the topic.

3. The method of claim 1, the assessing an emotional sentiment related to the first article comprising:

assessing an emotional sentiment related to the first article as a function of a specified number of terms adjacent the first link.

4. The method of claim 1, comprising:

replacing the first link with a placeholder.

5. The method of claim 1, comprising:

replacing the first link with the second link.

6. The method of claim 1, the second article comprising content similar to the first article when a fraction of words in W(a_i) that are covered by at least one set of term sequences that occur in both the first article a_i and the second article a_j exceeds a predetermined threshold, where W(a_i) is a substring of all terms in the first article a_i not covered by frequent terms.

7. The method of claim 1, comprising acquiring the first social media content from a syndicated feed crawler that crawls a social media platform.

8. The method of claim 1, the first social media content comprising at least one of a blog posting, microblog posting, weblog posting, usenet posting, or message board posting.

9. The method of claim 1, the first article comprising a news article.

10. A system for assessing an emotional sentiment related to a topic, comprising:

one or more processing units; and

memory comprising instructions that when executed by at least some of the one or more processing units perform operations, comprising;

identifying first social media content comprising a first link to a first article associated with the topic;

assessing an emotional sentiment related to the first article as a function of one or more terms in the first social media content;

identifying second social media content comprising a second link to a second article associated with the topic and comprising content similar to the first article, the second article comprising content similar to the first article when S(a_i,a_j)/size(a_i) exceeds a predetermined threshold, where S(a_i,a_j) is a number of terms of a largest set of infrequent terms between the first article (a_i) and the second article (a_j) and size (a_i) is the number of terms in the larger of articles a_i and a_j;

assessing an emotional sentiment related to the second article as a function of one or more terms in the second social media content; and

aggregating the emotional sentiment related to the first article with the emotional sentiment related to the second article to assess the emotional sentiment related to the topic.

11. The system of claim 10, the operations comprising:

replacing the first link with the second link.

12. The system of claim 10, the second article comprising content similar to the first article when a fraction of words in W(a_i) that are covered by at least one set of term sequences that occur in both the first article a_i and the second article a_j exceeds a predetermined threshold, where W(a_i) is a substring of all terms in the first article a_i not covered by frequent terms.

13. A computer readable storage device comprising computer executable instructions that when executed perform a method for assessing an emotional sentiment related to a topic, comprising:

identifying first social media content comprising a first link to a first article associated with the topic;

assessing an emotional sentiment related to the first article as a function of one or more terms in the first social media content;

identifying second social media content comprising a second link to a second article associated with the topic and comprising content similar to the first article, the second article comprising content similar to the first article when S(a_i,a_j)/size(a_i) exceeds a predetermined threshold, where S(a_i,a_j) is a number of terms of a largest set of infrequent terms between the first article (a_i) and the second article (a_j) and size (a_i) is the number of terms in the larger of articles a_i and a_j;

assessing an emotional sentiment related to the second article as a function of one or more terms in the second social media content; and

aggregating the emotional sentiment related to the first article with the emotional sentiment related to the second article to assess the emotional sentiment related to the topic.

14. The computer readable storage device of claim 13, the assessing an emotional sentiment related to the first article comprising:

assessing the emotional sentiment related to the first article via a maximum entropy classifier configured to output a binary output indicative of the emotional sentiment related to the first article and a confidence score indicative of a confidence of the binary output.

15. The computer readable storage device of claim 13, the first social media content comprising at least one of a blog posting, microblog posting, weblog posting, usenet posting, or message board posting.

16. The computer readable storage device of claim 13, the method comprising:

replacing the first link with the second link.

17. The computer readable storage device of claim 13, the second article comprising content similar to the first article when a fraction of words in W(a_i) that are covered by at least one set of term sequences that occur in both the first article a_i and the second article a_j exceeds a predetermined threshold, where W(a_i) is a substring of all terms in the first article a_i not covered by frequent terms.

18. The computer readable storage device of claim 13, the method comprising providing for display in a user interface at least one of:

a first graphical representation of the emotional sentiment related to the first article;

a second graphical representation of the emotional sentiment related to the second article;

or

a third graphical representation of the emotional sentiment related to the topic.

19. The computer readable storage device of claim 13, the method comprising:

providing for display in a user interface a graphical representation of the emotional sentiment related to the topic;

detecting user interaction with the graphical representation; and

displaying at least one of the first social media content, the second social media content, the first article, or the second article responsive to the detecting.

20. The computer readable storage device of claim 13, the method comprising:

acquiring the first social media content from a syndicated feed crawler that crawls a social media platform.