关键词: 不完备日志, 联合发生关系, 变化挖掘, 行为轮廓, ProM仿真

Abstract: To solve the problem of mining business system behavior changes from logs,a mining method based on incomplete logs and conjoint occurrence relations of logs was proposed.When the original reference model of business system was unknown,the incomplete logs without hidden transitions were used to obtain the conjoint occurrence relations of activities in the logs.By extracting the occurrence invariable sets of the activities,the delete,insert and move change operations (denoted as delete,Insert and Move respectively) in the logs were mined,and the method of behavior change mining driven by the logs was realized.ProM simulation verified that the proposed method could realize the log mining of system behavior changes.The experimental results showed the effectiveness and correctness of the proposed method.

Key words: incomplete logs, conjoint occurrence relation, change mining, behavioral profile, ProM simulation