BPDetector:一种可扩展的业务模式识别框架  

BPDetector:extensible business pattern detector framework

作　　者：闾亚 雷建坤[1,2] 杨丽琴[1,2] 张亮[1,2] 

机构地区：[1]复旦大学计算机科学技术学院,上海201203 [2]复旦大学上海市数据科学重点实验室,上海201203

出　　处：《计算机集成制造系统》2016年第2期287-293,共7页Computer Integrated Manufacturing Systems

基　　金：国家自然科学基金资助项目(60873115);教育部-中国移动科研基金资助项目(MCM20123011);上海市科技发展基金资助项目(13dz2260200;13511504300);国家863计划重大资助项目(2012AA02A602)~~

摘　　要：为了有效地发现业务模式,提出一种基于复杂事件处理技术的可扩展业务模式识别框架,该框架不仅可以从日志中发现业务模式以及相应的配置参数,还具可伸缩特征,即框架与待发现的业务模式及其个数不存在紧密关系。以中国移动公司办公系统内部的业务流程执行日志为例,基于Esper复杂事件处理开源工具开发了业务模式识别系统,验证了该方法的可行性和有效性。To leverage the business process consolidation,an extensible framework for business pattern discovering based on Complex Event Processing（CEP）technique was proposed.The framework could detect the business pattern and configuration parameters from log as desired,and distinguish itself from conventional methods by the scalability,in which the framework complexity was decoupled from the number of patterns.The implementation of the framework based on an open source CEP tool named Esper was elaborated.An empirical study was conducted on the business process logs from the office automation system in China Mobile Communication Corporation（CMCC）to justify the feasibility and effectiveness of the proposed approach.

关 键 词：业务模型 业务模式识别 可扩展框架 复杂事件处理 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]