Behavior analysis and event detection with statistics of news release on insurgent website
Min Tang, Duoyong Sun, Bo Li, Zihan Lin · 2016
Behavior analysis and events detection of insurgent organization are important issues in intelligence information research and public security management. Due to the covertness of insurgent activities, the relationships between organization and events are hard to detect. Using open source data provides an effective way to obtain information of organizational activities. One important issue in this field is mining hidden frequency pattern with noisy or incomplete information. In this paper, we study the activities of insurgent organization and related event with news data mined from insurgent website. A covert behavior mining method is proposed to process the open source data and analyze news release behaviors. The method can mine frequency pattern of news releasing in the case of losing part of news information. Three methods are employed to detect the events based on the processed data. The performance of the proposed method is evaluated through real data. With the results, we show how this method can be used to improve the events detection and surveillance of insurgent organizational activities.