基于用户画像与关联规则的图书馆资源组合推荐算法

Library Resource Combination Recommendation Algorithm Based on User Portrait and Association Rules

中文关键词: 用户画像；关联规则；图书馆资源；资源组合推荐算法；词汇相似度

英文关键词: user portrait, association rules, library resources, resource combination recommendation algorithm, lexical similarity

中文摘要:

      图书馆资源推荐算法难以与用户的偏好相结合，导致其推荐精度较差。为提高推荐结果的准确性，基于用户画像与关联规则设计图书馆资源组合推荐算法。通过用户画像算法得到图书资源兴趣估计，在特征样本集合的基础上，计算主题权重的标准值，建立用户情景兴趣度表达式。基于关联规则建立图书馆资源聚合模型，计算文档内相同词汇出现的频率，计算不同书籍的相似度，并在书籍指标权重的基础上，得到关联规则下图书馆资源的聚合函数，以此设计资源组合推荐算法。实验结果表明，其最高精确率、召回率、F1值分别为0.92、0.73 和0.69， 该推荐算法的推荐精度较高。

英文摘要:

      Library resource recommendation algorithm is difficult to combine with users’ preferences, resulting in poor recommendation accuracy. In order to improve the accuracy of recommendation results, library resource combination recommendation algorithm is designed based on user portrait and association rules.The interest estimation of book resources is obtained through the user portrait algorithm. Based on the feature sample set, the standard value of subject weight is calculated, and the expression of user scenario interest is established. The library resource aggregation model is established based on association rules, the frequency of the same vocabulary in the document, and the similarity of different books are calculated. Based on the book index weight, the aggregation function of library resources under association rules is obtained, so as to design the resource combination recommendation algorithm. The experimental results show that its highest accuracy,recall, and F1 values are 0.92, 0.73, and 0.69, respectively, indicating that the recommendation accuracy of this recommendation algorithm is superior to other algorithms.