An Information Retrieval System with Ability of Analogical Inference Using Semantic Network and Function of Fuzzification
暂无分享,去创建一个
Abstract In information retrieval system, it is necessary to grasp user's subject of interest in order to present appropriate documents to the user. Tn this paper, we propose a model of human ability of analogical inference based on association between key words and, using it, construct an information retrieval system in which the computer with the ability learns its user's subject of interest through question-answering with the user. In this system, the association between key words is represented by a semantic network , and a function of fuzzification of input information is introduced in the semantic network to implement the ability of analogical inference based on the association. Finally, the effect of analogical inference on the learning efficiency of the system is investigated.
[1] Lauren B. Doyle,et al. Indexing and abstracting by association , 1962 .
[2] Edward H. Shortliffe,et al. Production Rules as a Representation for a Knowledge-Based Consultation Program , 1977, Artif. Intell..
[3] Robert Fugmann,et al. Representation of concept relations using the TOSAR system of the IDC: Treatise III on information retrieval theory , 1974, J. Am. Soc. Inf. Sci..