Font Size: 

With the rapid advancement of social networking services, people tend to exchange and share information online. Massive global information are aggregated promptly and circulated quickly via the social networks, such as Facebook, LINE, PTT and Dcard. How to extract useful information from the social networks to support decision making on public events is one of the most important research issues for the Taiwan government. In this paper, we focus on the information integration of disaster events and the analysis of public sentiment on social networks. Three subjects are thoroughly investigated: (1) predictive analysis on disaster information in social networks via natural language processing and semantic model analysis; (2) information extraction and sentiment trend analysis for disaster events on social networks and (3) a crowdsourcing based correctness verification approach of information on social networks.

Crowdsourcing; Disaster Prevention; Refugee Rescue; Public Sentiments Analysis; Social Network; Semantic Model Analysis