基于描述逻辑方法的VSM语义检索模型

DL-VSM:Description logic-based VSM for semantic retrieval

针对当前的信息检索模型并不能提供语义信息的检索问题,提出了一个基于描述逻辑方法的语义检索模型,定义了文档的逻辑视图、查询的逻辑视图和两种视图之间的相似度计算方法,并给出了模型的存储结构。该模型将用户的检索请求和待查询的数据(文档)转化成基于描述逻辑知识库为基础的个体集合,不仅能够有效表示文档和查询的语义信息,而且有利于计算机自动推理的实现,可以有效提高检索的准确率和召回率。

Aimed at the problem that current IR model could not supply the sematic retrieval,DL-VSM（description logic-based VSM for semantic retrieval） is proposed.This model expressed the documents and user＇s requests by the ontology descripted by description logic and put forward the method which can compute the similarity between the logic view of documents and requests.The data structure of this model is given.Experiment and model analyse show that F-value is improved 10% or so higher than VSM and this model performed very well.