一种基于类比的启发式搜索方法

A Heuristic Search Method by Analogy

引言状态空间的启发式搜索一直是Al领域最主要的问题求解方法,过去己取得了不少成果,如A‘算法和AO“算法等。但是在实际应用中,这些启发式搜索方法所使用的启发信息大多是人们依据具体领域问题靠经验总结得来的,启发信息的获取十分困难,且其精确性和可靠性也难以保证。另一方面,目前的状态空间搜索方法大多是一次性搜索,将同一问题状态空间中的多次搜索视为彼此独立。

Integrating analogical reasoning into the heuristic method in state-space search, we present a new heuristic search method by analogy. It can reduce the search effort and solve problems more effectively. This paper describes the idea and the strategy of the search method, discusses the solving model which includes three reasoning processes: generating a case, retrieving a case and guiding the search for a solution.