图书信息检索中读者兴趣偏好挖掘模型的建立仿真

Establishment and simulation of mining model for interest preference of readers in book information retrieval

针对传统挖掘模型对图书信息检索中读者兴趣偏好数据进行挖掘时,存在的挖掘效率低、耗时长等问题,提出基于聚类分析的读者兴趣偏好挖掘模型.采用分类索引分布树法对图书相似度与读者兴趣距离进行计算,通过兴趣因子对兴趣偏好度进行度量,并以此为基础,采用相关反馈模型Rocchio算法对读者兴趣图书检索进行扩展,引入聚类分析法建立图书信息检索中读者兴趣偏好挖掘模型.实验仿真结果表明,采用改进模型时,其挖掘效率、时间及误差均优于传统模型.

Aiming at the problem that such detects as lowmining efficiency and large error always exist when the traditional mining model is used to mine the interest preference data of readers in the book information retrieval,a mining model for the interest preference of readers was proposed. The distance between the book similarity and reader interest was calculated with the classification index distribution tree method. In addition,the interest preference level was measured through interest factor. On this basis,the interest book retrieval for readers was extended with the relevance feedback model Rocchio algorithm,and the mining model for interest preference of readers in the book information retrieval was established with the clustering analysis method. The results showthat the mining efficiency,time and error of the improved model are superior to those of the traditional models.