By Vedant, Ashwin Kumar, Abhishek Ranjan, Mayank Srivastava
Team White Walkers
WHY? -- Information in news is highly scattered over space and time. So it becomes difficult to see everything in news, relate all incidents and draw out conclusions. WHAT? -- We give an interface where user can search the name of an entity* in news and he gets a graph of the entities which are related to that entity upto a certain number of hops. *entity: Any proper noun which has significance in news. They are related to other entities through well-defined relations. HOW ? -- We have downloaded past archives from some newspapers. Also, our database is continuously growing with the RSS feeds from those newspapers. We then send those news articles to an NLP tool which extracts the entities and relations from it. After that, we generate a graph out of them and based on the user's query, we display him the desired information. We also provide some additional information about nodes which are extracted from some structured sources.