隨著信息技術與互聯(lián)網(wǎng)的飛速發(fā)展,高校讀者僅靠傳統(tǒng)的基于檢索的服務很難從海量圖書中發(fā)現(xiàn)真正感興趣的或對其有價值的圖書。應用數(shù)據(jù)挖掘技術和個性化推薦技術,根據(jù)讀者自身信息需求的差異,將符合讀者需求的圖書主動推薦給讀者,這種主動服務方式不僅提高了高校圖書館的服務水平,使高校圖書館發(fā)展地更加全面、人性化,還可以發(fā)掘讀者潛在的信息需求,提高館藏圖書的借閱率,將圖書資源利用率最大化。論文旨在研究基于聚類算法的高校圖書館個性化推薦的一般方法,利用常見的幾類推薦算法,設計適用于高校圖書館的個性化推薦策略,并將華中師范大學2017年全年圖書館圖書借閱信息及2014級-2017級本科生個人信息作為實例數(shù)據(jù),進行數(shù)據(jù)挖掘,從而驗證前面設計的個性化推薦策略的可行性。論文首先建立讀者-類目偏好模型,對目標讀者基于這一偏好模型進行聚類分析。在聚類分析的過程中,首先將原始的k-means聚類算法優(yōu)化,再使用改進后的K-Means算法實現(xiàn)讀者的聚類。接下來,將協(xié)同過濾推薦算法與基于內容的推薦算法結合,起到同時實現(xiàn)兩個推薦算法的優(yōu)點的目的。利用協(xié)同過濾和基于內容的混合推薦(基于興趣的推薦)策略,針對每個讀者進行圖書的個...

【文章頁數(shù)】：112 頁

【學位級別】：碩士

【文章目錄】：
Abstract
摘要
Acknowledgements
Chapter 1 Introduction
    1.1 Research Background and Significance
        1.1.1 Research Problem
        1.1.2 Research Significance
    1.2 Related Work
    1.3 Research Content and Method
        1.3.1 Research Content
        1.3.2 Research Method
    1.4 Thesis Structure
Chapter 2 Theoretical Basis
    2.1 Theoretical Basis of Cluster Analysis
        2.1.1 Concept of Reader Segmentation
        2.1.2 Introduction of Data Mining
        2.1.3 Cluster Analysis
        2.1.4 Basic K-Means Algorithm
    2.2 Theoretical Basis of Personalized Recommendation in University Library
        2.2.1 Overview of Recommendation System
        2.2.2 Recommendation Method Used in University Library
        2.2.3 Comparison of 3 Recommendation Algorithms
    2.3 Introduction of Chinese Library Classification
    2.4 Summary
Chapter 3 Related Process of Personalized Recommendation for University Library
    3.1 Optimization Methods on Basic K-Means Algorithm
        3.1.1 Description of Optimized K-Means Algorithm
        3.1.2 Algorithm Characteristics
        3.1.3 Performance Evaluation of Optimized K-Means Algorithm
    3.2 Design the Framework of Personalized Recommendation in University Li-braries
        3.2.1 "My Library" in CCNU
        3.2.2 Demand Analysis of Personalized Recommendation System in CCNULibrary
        3.2.3 Framework of Personalized Recommendation in CCNU Library
    3.3 Reader Preference Model Based on Classification Number in CLC
    3.4 Algorithm and Strategy of Personalized Recommendation in University Li-brary
        3.4.1 Introduction of Collaborative Filtering Recommendation Algorithm
        3.4.2 Introduction of Content-Based Recommendation Algorithm
        3.4.3 Introduction of Collaborative Filtering and Content-Based CombinedRecommendation Strategy
    3.5 Summary
Chapter 4 Experimental Settings and Results Analysis
    4.1 Data Collection and Preprocessing
        4.1.1 Selection of Data Sources
        4.1.2 Data Cleaning
        4.1.3 Data Conversion
        4.1.4 Data Integration
    4.2 Cluster Algorithm on Readers
        4.2.1 Simulation Experiment on Partial Readers
        4.2.2 Experiment on All Readers
    4.3 Personalized Recommendation on Readers
        4.3.1 Personalized Recommendation Booklist Based on Readers' Interests
        4.3.2 Personalized Recommendation Booklist Based on Faculty Ranking
    4.4 Results Analysis and Evaluation
        4.4.1 Evaluation Indicators for Personalized Recommendation
        4.4.2 Evaluation of Personalized Recommendation of CCNU Library
    4.5 Summary
Chapter 5 Conclusion and Future Work
    5.1 Conclusion
    5.2 Future Work
References



本文編號：3797904