Statistical Analysis & Categorization for Tweets during Natural Disaster using Classification and Ranking Approach

Micro blogging sites are gaining an active role in the field of managing disaster. It is a blog that lets users publish short text updates. However, our work is based on information retrieval i.e., recovery of specific data from a huge dataset. Social media plays a crucial role in today's world, thus if in some part of the world, any disasters have occurred, it becomes sensational and people start tweeting about the disaster in the social media and make others aware about their safety. Therefore, it becomes imperative that we verify this information in order to gather relevant details which we refer to as fact-checkable tweets. In this paper, we have used the Nepal earthquake Dataset to determine the fact-checkable tweets from the non-fact-checkable once. We have generated a hybrid algorithm which has incorporated both classification and ranking approach. This methodology thus provides us with better outcome which outperforms other major previously proposed methodologies.