【摘要】 当前的书籍信息资源自动检索系统往往直接采用挖掘算法对信息进行检索,对共现词语的分析不足,因此在需要高查全率的检索中,存在查准率不足的问题。因此文中提出基于读者阅读偏好的书籍信息资源自动检索系统设计。硬件设计中设立异构多核的信息检索硬件结构。在软件设计中,首先根据我国中图分类规则,建立起分类树状结构,同时利用MFP-Miner算法对读者的阅读偏好进行运算,对偏好以及中图分类树状结构中出现的词语共现情况进行概念扩展根据此来进行检索。为了验证设计的系统是否满足设计初衷,使用所提系统及传统检索系统对某图书馆中的书籍信息进行检索。实验结果显示,文中设计的书籍信息资源自动检索系统的查准率在查全率提高的情况下,降低程度较低,满足设计初衷。更多还原

【Abstract】 In the current automatic retrieval system of book information resources,the mining algorithm is often used to retrieve information directly,and the analysis of co-occurrence words is insufficient,so that the precision is insufficient in the retrieval which needs high recall rate. Therefore,the design of book information resources automatic retrieval system based on readers’ reading preference is put forward in this paper. In the hardware design,The hardware structure of heterogeneous multicore information retrieval is set up,In the software design,the classification tree structure is established according to the classification rules in “Chinese Library Classification”, MFP miner algorithm is used to calculate the readers’ reading preference,and a concept expansion of the preference and the co-occurrence of words appearing in the tree structure of Chinese library classification is conducted for the book information resource retrieval. In order to verify whether the designed system meets the original intention of the design,the experiment was designed. The book information in a library was retrieved by using the proposed system and the traditional retrieval systems. The experimental results show that the accuracy ratio of the system reduces a little bit while check rate is increased,which can meet the original intention of the design.更多还原