Modeling social networks using analytic measurements of online social media content

• US 8,375,024 B2
• Filed: 01/13/2009
• Issued: 02/12/2013
• Est. Priority Date: 11/13/2008
• Status: Expired due to Fees

• Current Assignee: Buzzient, Inc.
• Original Assignee: Buzzient, Inc.
• Inventors: Goeldi, Andreas
• Primary Examiner(s): LE, DEBBIE M
• Application Number:
  US12/353,166
• Publication Number:
  US 20100121849A1
• Time in Patent Office: 1,491 Days
• Field of Search: 707/705, 707/706, 707/722, 707/736, 707/999.006, 707/7, 707/104.1
• US Class Current: 707/722
• CPC Class Codes: G06Q 10/00 Administration; Management
• CPC Class Codes: G06Q 30/00 Commerce
• CPC Class Codes: G06Q 30/0251 Targeted advertisements
• CPC Class Codes: G06Q 30/0263 based upon Internet or website rating
• CPC Class Codes: G06Q 30/0277 Online advertisement
• CPC Class Codes: G06Q 50/01 Social networking

Abstract

Methods, apparatuses, and computer-readable media for generating a social network graph to model one or more social networks of related authors of online social media and their corresponding posts of online social media conversations relevant to subject matter of interest in a category. Embodiments are configured to harvest and aggregate posts of online social media conversations from one or more online social media sources; to perform content scraping on the posts of online social media conversations to obtain raw data that includes user-profile information of a social media author corresponding to each individual post of online social media conversations; and performing social network analysis processing on the raw data to obtain the social network graph.

Claims

1. A method comprising:

generating a social network graph to model one or more social networks of related authors of online social media and their corresponding posts of online social media conversations relevant to subject matter of interest in a category, wherein generating the social network graph comprises;

harvesting and aggregating posts of online social media conversations from one or more online social media sources;

performing content scraping on the posts of online social media conversations to obtain raw data including each individual post of an online social media conversation relevant to the subject matter of interest and user-profile information of a social media author corresponding to each individual post; and

performing social network analysis processing on the raw data to obtain the social network graph, the processing including;

analyzing an influence level of the social media author, the influence level indicative of a total number of other authors the social media author is connected to over one or more social networking paradigms, a total number of posts written by the author in one or more social networking paradigms, an importance score associated with the one or more social networking paradigms in which the social media author'"'"'s posts are published, a total number of distinct subgroups of authors the social media author is connected with, and an overall activity level of the social media author;

computing an influence score associated with the social media author using an influence score formula, the influence score formula computed by multiplying a raw betweenness centrality value for the social media author with a function of;

(1) a number of other active authors on the website where the social media author is active, (2) a total number of posts the social media author has contributed, and (3) one or more correction parameters that are fine-tuned for purposes of a specific vertical that corresponds to the specific subject matter of interest;

wherein the betweenness centrality value is computed as a result of a centrality measure of a vertex within a raw social graph generated for the given social networking paradigm;

detecting a sentiment rating associated with a one or more posts of the social media author; and

determining weighting factors associated with the one or more posts based on a combination of the sentiment rating associated with the one or more posts and the influence level associated with the social media author.

2. The method of claim 1, further comprising storing the social network graph in a database.

3. The method of claim 1, wherein harvesting the online social media conversations includes:

performing forum analysis to discover websites of interest containing online social media content relevant to the subject matter of interest in the category; and

searching through the websites of interest to extract social media conversations related to the subject matter of interest.

4. The method of claim 3, wherein the forum analysis comprises:

searching for websites with online social media content relevant to the subject matter of interest using keywords in a search engine;

assigning a relevance score to each website found in search engine results based on the keywords; and

designating websites with a high relevance score as websites of interest.

5. The method of claim 1, wherein the content scraping comprises:

identifying hyperlinks to potentially relevant online social media conversations in one or more web pages of each of the websites of interest;

iteratively accessing and retrieving multiple levels of web pages within the each of the websites of interest using the hyperlinks until relevant conversation threads are found;

breaking down each relevant conversation thread into individual posts of online social media conversations; and

storing the raw data in the database.

6. The method of claim 5, wherein the raw data further comprises information on those social media authors responding to each of the individual posts.

7. The method of claim 1, wherein the user-profile information of the social media authors includes one or more of:

author'"'"'s username;

author'"'"'s demographic and geographic information;

number of conversations the author has posted to the website; and

online social media connections of the author.

8. The method of claim 1, wherein the social network analysis processing comprises:

determining influence of each of the authors within the one or more social networks of related authors of online social media and their corresponding posts;

ranking the influence of each of the authors to ascertain most influential authors for the category; and

compiling a list of the most influential authors for the category sorted by influence.

9. The method of claim 8, wherein the influence of each of the online social media authors is determined by:

calculating a centrality value for each author using a betweenness centrality algorithm on the raw data stored in the database for each of the websites of interest, the centrality value to indicate a degree of influence of the author in the one or more social networks; and

calculating an influence score for each author based on the centrality value of the author to determine the influence of each of the authors within the one or more social networks of related authors of online social media and their corresponding posts.

10. The method of claim 9, wherein the influence score for each of the authors is further based on a number of active authors on the websites of interest and number of posts to the websites of interest by each author.

11. The method of claim 9, further comprising modifying the centrality value by an activity level of each of the authors and by an importance of the website where the author is active.

12. The method of claim 1, further comprising:

parsing the social network graph to obtain early warnings about a service issue or quality issue associated with a particular service or product that embodies the subject matter of interest.

13. The method of claim 1, further comprising:

identifying a first list of most influential authors corresponding to a particular subject matter of interest based on a ranking of the influence levels associated with a plurality of social media authors participating in the online social media conversations.

14. The method of claim 1, further comprising:

based on the social network graph, generating a list of advertising networks that place advertisements that place advertisements in the analyzed online social media;

utilizing the social network graph, identifying advertising networks that place advertisements most relevant to the subject matter of interest in the analyzed social media; and

outputting the identified list of relevant advertising networks to a user of the social network graph.

22. The system of claim 1, wherein the social network analysis processing comprises:

determining influence of each of the authors within the one or more social networks of related authors of online social media and their corresponding posts;

ranking the influence of each of the authors to ascertain most influential authors for the category; and

compiling a list of the most influential authors for the category sorted by influence.

15. A system comprising:

a processor;

memory coupled to the processor, the memory configured to store a set of instructions corresponding to a method executed by the processor, the method including;

generating a social network graph to model one or more social networks of related authors of online social media and their corresponding posts of online social media conversations relevant to subject matter of interest in a category, wherein generating the social network graph comprises;

harvesting and aggregating posts of online social media conversations from one or more online social media sources;

performing content scraping on the posts of online social media conversations to obtain raw data including each individual post of an online social media conversation relevant to the subject matter of interest and user-profile information of a social media author corresponding to each individual post; and

performing social network analysis processing on the raw data to obtain the social network graph, the processing including;

analyzing an influence level associated with the social media author, the influence level indicative of a total number of other authors the social media author is connected to over one or more social networking paradigms, a total number of posts written by the author in one or more social networking paradigms, an importance score associated with the one or more social networking paradigms in which the social media author'"'"'s posts are published, a total number of distinct subgroups of authors the social media author is connected with, and an overall activity level of the social media author;

computing an influence score associated with the social media author using an influence score formula, the influence score formula computed by multiplying a raw betweenness centrality value for the social media author with a function of;

(1) a number of other active authors on the website where the social media author is active, (2) a total number of posts the social media author has contributed, and (3) one or more correction parameters that are fine-tuned for purposes of a specific vertical that corresponds to the specific subject matter of interest;

wherein the betweenness centrality value is computed as a result of a centrality measure of a vertex within a raw social graph generated for the given social networking paradigm;

detecting a sentiment rating associated with a one or more posts of the social media author; and

determining weighting factors associated with the one or more posts based on a combination of the sentiment rating associated with the one or more posts and the influence level associated with the social media author.

16. The system of claim 15, wherein the method further comprises storing the social network graph in a database.

17. The system of claim 15, wherein harvesting the online social media conversations includes:

performing forum analysis to discover websites of interest containing online social media content relevant to the subject matter of interest in the category; and

searching through the websites of interest to extract social media conversations related to the subject matter of interest.

18. The system of claim 17, wherein the forum analysis comprises:

searching for websites with online social media content relevant to the subject matter of interest using keywords in a search engine;

assigning a relevance score to each website found in search engine results based on the keywords; and

designating websites with a high relevance score as websites of interest.

19. The system of claim 15, wherein the content scraping comprises:

identifying hyperlinks to potentially relevant online social media conversations in one or more web pages of each of the websites of interest;

iteratively accessing and retrieving multiple levels of web pages within the each of the websites of interest using the hyperlinks until relevant conversation threads are found;

breaking down each relevant conversation thread into individual posts of online social media conversations; and

storing the raw data in the database.

20. The system of claim 19, wherein the raw data further comprises information on those social media authors responding to each of the individual posts.

21. The system of claim 15, wherein the user-profile information of the social media authors includes one or more of:

author'"'"'s username;

author'"'"'s demographic and geographic information;

number of conversations the author has posted to the website; and

online social media connections of the author.

23. The system of claim 15, wherein the method further comprises:

parsing the social network graph to obtain early warnings about a service issue or quality issue associated with a particular service or product that embodies the subject matter of interest.

24. The system of claim 15, wherein the method further comprises:

identifying a first list of most influential authors corresponding to a particular subject matter of interest based on a ranking of the influence levels associated with a plurality of social media authors participating in the online social media conversations.

25. The system of claim 15, wherein the method further comprises:

based on the social network graph, generating a list of advertising networks that place advertisements that place advertisements in the analyzed online social media;

utilizing the social network graph, identifying advertising networks that place advertisements most relevant to the subject matter of interest in the analyzed social media; and

outputting the identified list of relevant advertising networks to a user of the social network graph.