GEOGRAPHICAL PRIORITIZATION OF SOCIAL NETWORK MESSAGES IN NEAR REAL-TIME USING SENSOR DATA STREAMS: AN APPLICATION TO FLOODS

Abstract

Social networks have been used to overcome the problem of incomplete offi cial data, and to provide a more detailed description of a disaster. However, the fi ltering of relevant messages on-the-fl y remains challenging due to the large amount of misleading, outdated or inaccurate information. This paper presents an approach for the automated geographic prioritization of social network messages for fl ood risk management based on sensor data streams. It was evaluated using data from Twitter and a monitoring agency of Brazil. The results revealed that the proposed approach is useful for identifying valuable fl ood-related messages in near real-time.