Sentiflood: Process model for flood disaster sentiment analysis

The growing utilization of Web 2.0 leads us to extract, transform, load, and analyze enormously and sizably voluminous amount of structured and unstructured data, at a speedy pace, mentioned to as ‘Big Data’. With the help of collected public opinion from social media, users are transforming themselves into a social sensor. Data produced by social media is believed can be important in understanding the public's reactions and feelings. Particularly for disaster management, finding posts that indicate a situation of dissatisfaction, danger or worrying may prove critical. Consequently, a systematic classification is genuinely helpful in processing these posts and classify them into sentiment polarity and aspect-based classification that will benefit diverse agencies such as non-government or government in managing such crisis situations. However, there is less work of other researchers in developing big data application using a systematic method such as methodology. Distinctly, in disaster management system that exploits sentiment analysis. Based on ATHENA project, this work extends the Crisis Information Processing Centre component by using supervised learning technique of machine learning approach with the incorporation of RUP/SOMA methodology.