Abstract
Owing to the widespread of the Internet, it is easier to reach information through the Internet. When performing environment scanning, organizations typically deal with a numerous of episodes and events about their core business, relevant technique standards, competitors, and market, among many others, where each episode or event to monitor or track generally is associated with many news documents. To reduce such information overload and information fatigues when monitoring or tracking events, it is essential to develop an effective episode evolution discovery technique to organize all news documents pertaining to an event of interest into an episode evolution graph. In this thesis, we propose a new feature selection metric, referred to as TF□□2 and develop an episode evolution discovery technique that uses the TF□□2 metric as its feature selection method and TF□IDF as the document representation scheme. Using the traditional TF□IDF as the performance benchmark, our empirical evaluation results suggest that our proposed TF□□2 technique outperforms its benchmarks and demonstrates the utility of TF□□2 metric for discovering episode evolution relationships.