Hotspot Detection and Analysis in Chat Environment

In the recent age of internet, one of the media that is quite commonly used by many people is social media which has become an important information resource for showing the happenings around the globe. A fast development in the social media like discussion forums, e-mail and chat environment resources are contributing greatly to the collective knowledge that remains unused. The instant communication is effectively attracting by many people in the chat environment. The sparseness of those data is a challenging for analysis. Therefore, it is necessary to understand the opinion of the users by analyzing the conversations. Determination of chat conversation topic is one of the important areas in chat mining. The conversation of the chat data is stored in the log files and that plays major role for performing the analysis. The proposed system aims to identify which topic or incident is more predominant in the world of cricket for a particular period of time. In order to identify the possible topics of the chat, the cricket news and posts collected from cricket sites and are grouped by clustering. The matching between chat log data and cricket site data is performed, where centroid of the cluster helps to decide the hot topic in the corresponding time window. The experiment on real data set provides meaningful information and significantly very much useful.