Searching for patterns in political event sequences: Experiments with the KEDS database

This paper presents an empirical study on the possibility of discovering in teresting event sequences and sequential rules in a large database of inter national political events. A data mining algorithm first presented by Manni la and Toivonen (1996), has been implemented and extended, which is able to search for generalized episodes in such event databases. Experiments condu cted with this algorithm on the Kansas Event Data System (REDS) database, a n event data set covering interactions between countries in the Persian Gul f region, are described. Some qualitative and quantitative results are repo rted, and experiences with strategies for reducing the problem complexity a nd focusing on the search on interesting subsets of events are described.