【摘要】 为了解决目前传统的信息检索工具返回大量无关的信息或漏检有用信息的问题,首先引入了本体的基本概念及其在信息检索中的用途,在此基础上提出了一种基于本体的智能信息检索系统模型。该模型首先使用基于SOM神经网络和分层聚类的两阶聚类算法自动的产生本体,免除了人工构造本体的繁琐,然后利用本体中概念及概念之间明确的关系描述,将用户提出的检索要求进行语义上的扩充,使信息检索过程更加智能化,大大提高了检索的查全率和查准率。更多还原

【Abstract】 At present,things that too much irrespective information returns or useful information is missed exist in the conventional information retrieval tools,they can’t meet user’s requirements.To solve this problem,the basic concepts of ontology and its application in information retrieval is introduced,subsequently an intelligent information retrieval system model based on Ontology is put forward.This system model first generates ontologies automatically by using a two-stage clustering approach based on SOM neural network and agglomerative hierarchical clustering,which avoids the people to construct ontologies manullay that are too time consuming and tedious,then utilizes the normalized concepts and the clear description of relations between the concepts in ontology to extend the retrieval requests in semantically which user put forward,thus intelligentizes the process of information retrieval and remarkably improves the accuracy rate and cover rate of retrieval.更多还原