Discovering topical experts in the Twitter social network
Description
Finding topical experts on online social sites with millions
of users, is an important and challenging problem. In this
project, we developed a novel methodology for discovering topical experts in Twitter.
We utilized the wisdom of the Twitter crowds to discover the topics of expertise
of individual users, and hence identified topical experts. Our methodology exploits
the Lists feature, using which a user can group Twitter users
who tend to tweet on a topic that is of interest to her, and follow
their collective tweets. Our key insight was that these Lists are often
carefully created by users to include experts, and the List meta-data
(names and descriptions) provide valuable semantic cues to their
topics of expertise. Thus, we inferred a user's expertise by
analyzing the meta-data of crowdsourced Lists that contain the user.
We mined List information to build Cognos,
a system for finding topical experts in Twitter. Evaluation based on a real-world deployment showed that Cognos
infers a user's expertise more accurately and comprehensively than
state-of-the-art systems that rely on the user's bio or tweet content.
In fact, Cognos yields results comparable to, if not better than, those given by
the official Twitter experts search engine for a wide range of queries.
Publications
Inferring Who-is-Who in the Twitter Social Network
Naveen Kumar Sharma, Saptarshi Ghosh, Fabricio Benevenuto, Niloy Ganguly, and Krishna P. Gummadi. ACM SIGCOMM Workshop on Online Social Networks (WOSN), Helsinki, Finland, August 2012.
[Highest rated paper at the workshop, included in ACM Computer Communication Review, volume 42, issue 4, pp. 533--538, October 2012]
Cognos: Crowdsourcing Search for Topic Experts in Microblogs
Saptarshi Ghosh, Naveen Kumar Sharma, Fabricio Benevenuto, Niloy Ganguly, and Krishna P. Gummadi. ACM SIGIR Conference (SIGIR), Portland, Oregon, USA, August 2012.
Services / Applications