图书馆海量藏书目录准确检索仿真

Simulation of Accurate Retrieval of Massive Books in Library

对图书馆海量藏书目录准确检索,可有效提高图书书目的查询效率。准确检索图书馆中海量藏书目录,需要计算模糊检索结果与用户查询的相关度,并利用相关度对检索结果排序,优化其概率分布,实现准确快速检索藏书书目。传统检索方法仅实现模糊书目检索,未考察结果与用户查询的相关度,且未对概率分布进行优化,存在检索精度低,速度慢的缺陷。提出基于JE分词的图书馆海量藏书目录准确检索方法。先计算出图书馆海量藏书目录检索质量和响应时间的评价值,采用JE分词方法对书目检索综合评价值进行分词实现模糊检索,计算出模糊检索结果与用户目标书目查询的相关度,利用相关度进行检索结果排序,采用重要抽样算法对检索结果排序状态进行概率分布优化,同时运用等分散抽样法在保证检索概率分布函数正确性评估指标不变的情况下,减小方差,完成藏书目录准确检索。实验结果表明,所提方法检索书目所用时间较短,检索结果与用户所需书目结果基本一致。

In this paper,we focus on a method to accurate retrieve massive library catalogue based on JE participle.Firstly,the retrieval quality of massive library catalogue and the evaluation value of response time were calculated,and then the method of JE participle was used to segment word of comprehensive evaluation value of bibliographic retrieval,so as to realize fuzzy retrieval.Moreover,the correlation degree between fuzzy retrieval result and objective bibliographic query was calculated and the retrieval results were sequenced by correlation degree.In addition,the importance sampling algorithm was used to optimize the probability distribution of ranking state of retrieval results.Meanwhile,the equal-scattered sampling method was used to reduce the variance and complete the accurate retrieval of bibliography under the condition of guaranteeing that the accuracy evaluation index of retrieval probability distribution function remained unchanged.Simulation results show that the proposed method spends a short time to retrieve the title catalogue,and the retrieval result is basically consistent with the title catalogue required by users.