A Temporal Frequent Itemset-Based Clustering Approach For Discovering Event Episodes From News Sequence

When performing environmental scanning, organizations typically deal with a numerous of events and topics about their core business, relevant technique standards, competitors, and market, where each event or topic to monitor or track generally is associated with many news documents. To reduce information overload and information fatigues when monitoring or tracking such events, it is essential to develop an effective event episode discovery mechanism for organizing all news documents pertaining to an event of interest. In this study, we propose the time-adjoining frequent itemset-based event-episode discovery (TAFIED) technique. Based on the frequent itemset-based hierarchical clustering (FIHC) approach, our proposed TAFIED further considers the temporal characteristic of news articles, including the burst, novelty, and temporal proximity of features in an event episode, when discovering event episodes from the sequence of news articles pertaining to a specific event. Using the traditional feature-based HAC, HAC with a time-decaying function (HAC+TD), and FIHC techniques as performance benchmarks, our empirical evaluation results suggest that the proposed TAFIED technique outperforms all evaluation benchmarks in cluster recall and cluster precision.