[4P2-GS-6] Web mining (2)

[4P2-GS-6-04] Event Popularity Analysis Introducing Network Structure on Social Media

〇Hiroshi Nagaya¹, Yukio Ohsawa¹, Teruaki Hayashi¹, Hiroyuki A. Torii², Kazuko Uno³ (1. Department of Systems Innovation, Graduate School of Engineering, The University of Tokyo, 2. Department of Chemistry, Graduate School of Science,The University of Tokyo, 3. Louis Pasteur Center for Medical Research)

Keywords:Topic detection, Social media, Text mining, Disaster situation

Event popularity analysis is an important task for grasping ongoing trends about some events from social media and the internet. Especially, it is required in a crisis situation from the viewpoint of appropriate information transmission and prevention of false rumor diffusion. In this paper, we propose Net-TF-SW: noise-robust and explainable topic popularity analysis method and apply it for tweets related to the Fukushima Daiichi Nuclear Disaster. We also compare with existing methods and show our method is more robust for noise than them.