针对用户需求复杂多变,提出了用户反馈2阶段Web服务查询方法。该方法引入基于Lucene规范的Web服务搜索引擎系统,该系统由WSDL处理模块、向量化模块、相似度计算模块和QoS模块组成,分别支持WSDL解析功能、WSDL索引建立与查询功能、自定义权重计算功能和QoS功能。在解析加拿大Guelph大学的2507个真实WSDL文档集合基础上,筛选出符合web服务功能标准的WSDL文档集,并对其进行服务索引建立和查询测试。实验结果表明,系统Web服务发现性能具有较高P@N值和NDCP值,验证了系统的可行性和有效性。
A two-phase Web service discovery method based on user feedback is proposed for the complex user requirements. Web service discovery system based on Lucene is introduced, it consists of WSDL processing module, the vectorization mod- ule, similarity calculation module and QoS module. The system supports the function of WSDL parsing, WSDL indexing and querying, custom weighting and QoS. Finally, with the real 2507 WSDL documents from Guelph University, a proper Web service discovery test collection is constructed to test the query performance of system. The experimental results demonstrate that the system has higher value of P@N and NDCP, which validates its feasibility and efficiency.