You have no items in your shopping cart.
ABSTRACT
The essence of on-line event detection and tracking is to find the first document that talks about an event inform the user about the new event and track the subsequent discussion of the event. In the present system used for online event detection and tracking the system is trained to deal with specific events whereby events (tweet streams) are detected filtered and classified into stream as either being relevant or irrelevant to the events of concern. Since the goal is to detect only certain types of events, corresponding set of keywords are used to filter the tweet stream and relevant classification is performed after the preprocessing step, since not all tweets contained relevant keyword. Also the current system is event specific. In this study, an online event detection and tracking was developed with PHP and MySQL to accept news streams, cluster and summarize the stream into different events groups. The summarization process include tracking of the said events. Test run shows that the output consist of new events and the subsequent follow up discussion organized in a chronological order. Also the issue of credibility of news medium was solved by using a credible news outfit, The Washington Post which is one of the oldest extant newspapers is also, one of the top 10 most widely read newspaper in the word
