Mining the change of events in environmental scanning for decision support

Abstract

An organization's environment is increasingly complex. Business demand on environmental scanning has significantly increased in recent years due to an attempt to assist management in planning an organization's strategies and responses. The conventional technique for environmental scanning is event detection from text documents such as news stories. Event detection methods recognize events while they neglect to discover the changes of events. This work develops an event change detection (ECD) approach that combines association rule mining and change mining techniques. Detecting changes of events aids managers in making fast responses to the change of external environments. Association rule mining is employed to discover the subject behaviors of events from news stories. Changes of events are identified by comparing the subject behaviors of events from different time periods. The discovered event changes can provide effective decision support for decision makers to capture environmental information in a timely manner and make adequate decisions.