Automatic detection of interlinked events for better disaster management
- 1 February 2014
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 595-600
Abstract
This paper exploits sentiment analysis based techniques to automatically identify the interlinked events from disaster related news coverage. Here the interlinked events include: (1) effect (loss/damage) due to disaster, (2) recovery effort applied by recovery agencies and (3) people's feedback related to the recovery effort. The main idea is to analyze the performance of the disaster recovery agencies through people's feedback with respect to the damage/loss occurred during the disaster. To automatically identify the interlinked events, we introduce the combined use of text mining based techniques with sentiment analysis based techniques. Finally, to automatically detect the impact of the recovery effort for better disaster management, we introduce the use of SentiWordNet and manually created vocabulary related to the disaster. Our experimental results show the effectiveness of the proposed system on automatically collected news stories. It is important to note that the proposed approach will be helpful in better disaster management and resource planning for the future.Keywords
This publication has 2 references indexed in Scilit:
- Twitter for crisis communication: lessons learned from Japan's tsunami disasterInternational Journal of Web Based Communities, 2011
- Twitter power: Tweets as electronic word of mouthJournal of the American Society for Information Science and Technology, 2009