CMRULES: An Efficient Algorithm for Mining Sequential Rules Common to Several Sequences

摘要: We propose CMRULES, an algorithm for mining sequential rules common to many sequences in sequence databases –not appearing frequently sequences. For this reason, the does not use a sliding-window approach. Instead, it first finds association prune search space items that occur jointly Then eliminates do meet minimum confidence and support thresholds according time ordering. evaluated performance of CMRULES three different ways. First, we provide analysis its complexity. Second, compared on public dataset with variation from literature. Results show is more efficient low thresholds, has better scalability. Lastly, report real application complex system.