[1E3] [General Session] 6. Web Mining

[1E3-01] Social issues extraction from social media using context information

〇Shuhei Kubota¹, Masanao Ochi¹, Ken Nagahama², Kanji Sakai², Takeshi Skaki¹, Junichiro Mori¹, Ichiro Sakata¹ (1. The University of Tokyo, 2. DENTSU PUBLIC RELATIONS INC.)

Keywords:social issues extraction

In recent years, the importance of making policy based on objective evidence has increased in order to make policy and evaluate efficiently in the field of policy. The purpose of this research is to extract social issues that people are interested in from social media.Futheremore, we propose a framework for overviewing social issues by analyazing the relationship between user community and social issues.Although the knowledge for extracting social issues effectively form noisy media where word and grammar has collapsed was few, it was shown that word dependency information is particulary effective in extracting social issues.