基于核化局部敏感哈希的快速文档检索方法  

Kernelized Local-Sensitive Hashing for Fast Document Retrieval

作　　者：詹增荣[1] 程丹[2] Zhan Zengrong;Cheng Dan(School of Information Engineering,Guangzhou Panyu Polytechnic,Guangzhou 511480,China;Department of Sport Industry,Guangzhou Polytechnic of Sport Vocational and Technical College,Guangzhou 510650,China)

机构地区：[1]广州番禺职业技术学院信息工程学院,广东广州510480 [2]广州体育职业技术学院体育产业系,广东广州510650

出　　处：《湖南科技大学学报（自然科学版）》2019年第3期75-83,共9页Journal of Hunan University of Science And Technology：Natural Science Edition

基　　金：广州市教育科学规划(201811675);广东省科技发展专项资金项目(706049150203);2018年度广东省普通高校特色创新类项目(2018GWTSCX057);广州市市属高校科研项目(1201620532)

摘　　要：在大规模文档中进行快速的相似搜索对绝大多数信息检索应用程序是至关重要的.基于局部敏感哈希的检索方法将高维的空间数据映射到低维的二进制海明空间,从而实现了快速搜索.本文给出了一个基于核化局部敏感哈希的快速文档检索方法,可以在保证时间效率下允许算法使用不同的相似函数进行快速检索.实验结果表明该方法在大规模文档集合检索中具有较好的效率和准确率.Fast similarity search for large databases is critical to most Information Retrieval(IR)applications.Local sensitive hash function was proved to be efficient,as it embed high-dimensional features into a low-dimensional Hamming space where items was fast searched.A kernelized local sensitive hash method for document retrieval was presented.The method made use of arbitrary kernel functions to perform fast search for a wide class of useful similarity function without losing time efficiency.The experiment result shows that this method has higher efficiency and better accuracy for large scale document retrieval.

关 键 词：局部敏感哈希 相似搜索 文档检索 核函数 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]