On Automatically Extracting Discoveries from User Generated Content

The internet is a rich source of knowledge that is constantly expanding as users generate content which needs to be organized and classified. Future versions of the web are likely to include intelligent software agents capable of using metadata to understand complex relationships and better support users in discovering new knowledge. Wikipedia is a rich source of knowledge, but with limited and inconsistent structure to its content there has been limited progress in developing agents to use this information. This paper investigates using natural language processing to add metadata to Wikipedia articles, and presents an application where users could use the metadata to discover an alternative destination.