TY - GEN
T1 - TweetSense
T2 - 7th ACM Web Science Conference, WebSci 2015
AU - Vijayakumar, Manikandan
AU - Umamaheshwar, Tejas Mallapura
AU - Kambhampati, Subbarao
AU - Talamadupula, Kartik
N1 - Funding Information:
We gratefully acknowledge the significant help from Sushovan De in this research. This research is supported in part by the ARO grant W911NF-13-1-0023, and the ONR grants N00014-13- 1-0176, N00014-13-1-0519 and N00014-15-1-2027, and a Google faculty research award.
Publisher Copyright:
© 2015 ACM.
PY - 2015/6/28
Y1 - 2015/6/28
N2 - As the popularity of Twitter, and the volume of tweets increased dramatically, hashtags have naturally evolved to become a de facto context providing/categorizing mechanism on Twitter. Despite their wide-spread adoption, fueled in part by hashtag recommendation systems, lay users continue to generate tweets without hashtags. When such "orphan" tweets show up in a (browsing) user's time-line, it is hard to make sense of their context. In this paper, we present a system called TweetSense which aims to rectify such orphan tweeets by recovering their context in terms of their missing hashtags. TweetSense enables this context recovery by using both the content and social network features of the orphan tweet. We characterize the context recovery problem, present the details of TweetSense and present a systematic evaluation of its effectiveness over a 7 million tweet corpus.
AB - As the popularity of Twitter, and the volume of tweets increased dramatically, hashtags have naturally evolved to become a de facto context providing/categorizing mechanism on Twitter. Despite their wide-spread adoption, fueled in part by hashtag recommendation systems, lay users continue to generate tweets without hashtags. When such "orphan" tweets show up in a (browsing) user's time-line, it is hard to make sense of their context. In this paper, we present a system called TweetSense which aims to rectify such orphan tweeets by recovering their context in terms of their missing hashtags. TweetSense enables this context recovery by using both the content and social network features of the orphan tweet. We characterize the context recovery problem, present the details of TweetSense and present a systematic evaluation of its effectiveness over a 7 million tweet corpus.
KW - Context
KW - Hashtags
KW - Rectification
KW - Regression model
KW - Social network
KW - Twitter
UR - http://www.scopus.com/inward/record.url?scp=84978123706&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84978123706&partnerID=8YFLogxK
U2 - 10.1145/2786451.2790157
DO - 10.1145/2786451.2790157
M3 - Conference contribution
AN - SCOPUS:84978123706
T3 - Proceedings of the 2015 ACM Web Science Conference
BT - Proceedings of the 2015 ACM Web Science Conference
PB - Association for Computing Machinery, Inc
Y2 - 28 June 2015 through 1 July 2015
ER -