基于改进的VSM的藏文期刊论文检索研究  

Research on Tibetan Periodical Papers Retrieval Based on Improved VSM

作　　者：仁青东主 头旦才让 尼玛扎西 Rinchen Dhondub;ThuptenTsering;Nyima Tashi(School of Information science and technology,Tibet University,Lhasa 850000,China;Computer College of Qinghai normal university,Key Laboratory of Tibetan information processing,Ministry of Education,Xining Qinghai 810008,China)

机构地区：[1]西藏大学信息科学技术学院,西藏拉萨850000 [2]藏文信息处理教育部重点实验室,青海西宁810008

出　　处：《计算机仿真》2020年第12期400-403,454,共5页Computer Simulation

基　　金：国家重点研发计划重点专项(2017YFB1402200);藏文信息处理教育部重点实验室项目(2018Z004)。

摘　　要：藏文期刊论文检索研究是目前藏文信息检索必须要解决的首要问题。在传统向量空间模型的检索方法的基础上,结合影响检索效率的位置特征项,提出了一种基于改进的VSM的藏文期刊论文检索方法,从而提高了检索质量与效率。实验结果表明,所提出的方法在查全率、查准率、综合评估率和效率方面都得到了一定的提高,与传统的向量空间模型相比,取得了良好的效果。The research on the retrieval of Tibetan journals is the primary problem that must be solved in Tibetan information retrieval.Based on the retrieval method of traditional vector space model and the location feature items af-fecting retrieval efficiency,a retrieval method based on improved VSM for Tibetan journals is proposed,which im-proves retrieval quality and efficiency.The experimental results show that the proposed method has a certain improve-ment in the recall rate,precision rate,comprehensive evaluation rate and efficiency,and has achieved good results compared with the traditional vector space model.

关 键 词：向量空间模型 藏文期刊论文 检索 相似度 排序 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]