一种可以挖掘隐含任务的算法研究  

Study on the Algorithm of Implicit Task Can be Mining

作　　者：王燊[1] 

机构地区：[1]北方民族大学,宁夏银川750021

出　　处：《无线互联科技》2016年第5期61-63,共3页Wireless Internet Technology

基　　金：北方民族大学研究生创新项目;项目编号:YCX1559

摘　　要：业务流程中存在的隐含任务在业务执行时并没有在日志中体现,这种任务主要服务于路由,不能被忽视。如果不能对它有效挖掘,可能影响最终的挖掘结果和模型的建立。文章对α算法进行改进,提出一种能发现并添加隐含任务的算法—。该算法参考任务间序关系矩阵确定并行任务;根据并行任务的位置分析了隐含任务可能出现的情况,向流程中添加隐含任务;然后合并相同隐含任务,去掉冗余隐含任务。文章用一个案例解释了算法的执行过程,通过实验验证了算法的可行性和有效性。最后分析了算法的局限性。The hidden tasks of business process, which is mainly serves for route, haven＇t been reflected in logs during the task execution, can not be ignored. It may effect the final results of mining and establishment of model, if the hidden tasks can＇t be mined availably. To improve the α-algorithm, a new α#-algorithm is proposed to find and insert the hidden tasks. Referring to the order relation matrix between tasks, the parallel tasks are determined. The α#-algorithm inserts hidden tasks in the workflow by studying the situations where a hidden task may appear based on the presences of parallel tasks, then merging the same hidden tasks and removing the redundant ones. In this thesis, the process of the algorithm is explained with a case. Its feasibility and effectiveness is demonstrated by experiments. Finally, we also analyze the limitations of α#-algorithm.

关 键 词：隐含任务 序关系矩阵 工作流挖掘 工作流网 PETRI网 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]