摘要
溯因推理是在给定的理论和观察下,求出对于该观察可能的解释.然而,在一般的溯因推理中,并不要求解释的直观性和归纳性.从日常生活和科学发现的某些规律出发,本文提出了同例和异例的概念,提出了一个基于同例和异例的溯因推理框架,并引入了可诱导和可允许这两个限定条件.接下来,本文给出在该框架下求出所有极大解释的算法AH并验证了它的可靠性和完备性.最后,一个具体的例子演示了算法AH.相对于一般的溯因推理系统,本框架主要有如下的优点:溯因得到的解释更符合直观,溯因得到的解释还可以用来推导出更多的新的语句.
Given a theory and observations, abduction tries to find explanations for observations, However,the present abduction reasoning mostly has few characteristics of induction. In daily life and scientific discovery, a new property Q of some individual was not explained arbitrarily. First,other individuals （similar cases）, which have the same property Q, would be found. Other properties of similar cases would be examined to find the possible cause of property Q. Second,other individuals （dissimilar cases） , which have not the property Q, would be found too. Other properties of dissimilar cases were not believed to be the cause of the property Q. Based on similar cases and dissimilar cases.the possible explanations for the observations could be found. In this paper, the notions of similar cases and dissimilar cases were presented in a formal method. Based on it, an abduction reasoning framework was introduced, For computing all maximal explanations,an algorithm was presented and its soundness and completeness were proven. Last, an example demonstrated this method. Compared with the present abductions, this framework has the following advantages：the framework has more characteristics of induction;the abducted explanations can be applied to other constants to generate more statements.
出处
《厦门大学学报（自然科学版）》
CAS
CSCD
北大核心
2008年第4期495-499,共5页
Journal of Xiamen University：Natural Science
关键词
溯因推理
逻辑程序
同例
异例
abduction
logic programming
similar cases
dissimilar cases
