基于模式的个性化服务定制方法

在服务互联网环境下，大量的用户需求可能被不同的用户同时提出。如何针对用户的个性化需求快速有效地构造服务解决方案是一个值得研究的问题。文中提出了一种基于模式的个性化服务定制方法，通过实验验证了文中所提算法的有效性。首先，利用历史服务请求及服务解决方案识别出的需求模式和服务模式，建立需求模式和服务模式之间的匹配关系。然后，提出一种基于模式的个性化服务定制算法（LPSC）来处理用户的个性化需求。在LPSC算法中，对于每个用户的个性化需求，根据用户需求的相似度进行分类构造虚拟需求。再用有限个数的需求模式去替代虚拟需求，通过需求模式与服务模式的匹配关系找到需求模式对应的最佳服务模式集，通过服务模式的组合来产生最终的服务解决方案。

A pattern-based personalized service customization method

In the service Internet environment, different requirements may be made simultaneously by different users. It is a great challenge to structure composite service quickly and effectively to fulfill personalized requirements.This paper presented a personalized service customization method based on patterns.The requirement patterns and service patterns from historical service request and service solutions were identified, and the matching relationship between the requirement patterns and service patterns were established. An algorithm called LPSC was proposed to deal with personalized requirements by which personalized requirements were classified according to the similarity between requirements and virtual requirements were constructed to replace them. Then each virtual requirement could be replaced by a limited number of requirement patterns.Through the matching relationship established between requirement patterns and service patterns,we could find optimal service pattern set that suited the requirement patterns. Finally, service solutions could be obtained by the composition of service patterns.Experimental results prove the effectiveness of the proposed algorithm.