摘要
提出了一个基于Mediator的Webservices无回溯反向链动态合成框架,在该框架下提供了一种产生式建模方法和一种用于消除合成语义冲突的参数级本体,然后使用一种无回溯反向链算法进行合成.该合成方法避免了对合成方案的优化,它的参数级本体在合成语义处理上也比前向链合成中的属性级本体的效率高,特别适合于对同样的用户输入和Webservices进行多次合成以生成相同的或者不同的输出.
Proposed in this paper is a non-backtrace backward chaining Web services dynamic composition framework based on mediator. Under this framework, a rule modeling method in composition and a parameter-level ontology for eliminating the composition semantic conflict are provided, and then a non-backtrace backward chaining algorithm is used to compose Web services. This composition approach avoids the optimization of composition plan, and its parameter-level ontology is more efficient in handling the semantic problems in composition than the attribute-level ontology in forward chaining composition approach and is suitable for composing many times to the same user inputs and the same Web services in order to generate the same outputs or different outputs.
出处
《计算机研究与发展》
EI
CSCD
北大核心
2005年第7期1153-1158,共6页
Journal of Computer Research and Development
基金
国家自然科学基金项目(60473124)
上海高校网格SECEInstitute
IBM中国研究中心基金项目