摘要
随着互联网络和信息技术的快速发展,人类获得信息的途径越来越多,然而如何在大量数据中获得符合用户喜好的信息,给用户带来更好的体验成为研究的重点,为此提出融合大数据挖掘的用户个性化POI推荐方法。首先通过对大数据推荐系统的分析,构建基于大数据用户个性化模型,并对操作平台的数据按照两层关联规则的方法进行数据挖掘,提高用户个性化网络数据的精度。然后生成LDA主题模型,采用分词处理和去停用词处理的方法对用户个性化输入数据进行预处理,利用困惑度Perplexity在拐点处的值作为评价指标衡量语言模型。最后通过JS距离公式作为衡量主题间匹配度的指标,实现POI与用户的匹配。实验结果表明,融合大数据挖掘的用户个性化POI推荐方法不仅具有较高的预测精度,还具有较高的覆盖率,能够为用户提供高质量的个性化推荐结果。
With the rapid development of the Internet and information technology, human beings have more and more ways to obtain information. However, how to obtain information in line with their preferences in a large amount of data information and bring a better experience to users has become the focus of research. Therefore, a user personalized POI recommendation method integrating big data mining is proposed. Firstly, through the simple analysis of the big data recommendation system, the user personalization model based on big data was constructed, and the data of the operation platform was mined according to the method of two-layer association rules to improve the user personalized network data. Then the LDA topic model was generated, the user personalized input data was preprocessed by word segmentation and de stop word processing, and the value of Perplexity at the inflection point was used as the evaluation index to measure the language model. Finally, the JS distance formula was used as an index to measure the matching degree between topics to realize the matching between POI and users. The experimental results show that the user personalized POI recommendation method integrating big data mining not only has high prediction accuracy, butalso has high coverage, and can provide users with high-quality personalized recommendation results.
作者
秦鹏
贾洪杰
霍兴瀛
邓朝艳
QIN Peng;JIA Hong-jie;HUO Xing-ying;DENG Zhao-yan(School of Mathematics and Computer Science,Liupanshui Normal University,Liupanshui Guizhou 553004,China;School of Computer Science and Co mmunication Engineering,Jiangsu University,Zhenjiang Jiangsu 212013,China;Library,Liupanshui Normal University,Liupanshui Guizhou 553004,China)
出处
《计算机仿真》
北大核心
2022年第6期355-358,385,共5页
Computer Simulation
基金
国家自然科学基金青年项目(61906077)
六盘水市科技计划项目(52020-2018-04-04)
贵州省教育厅基金项目(黔教合KY字[2019]125)
贵州省教育厅基金项目(黔教合XY字[2020]127)。
关键词
大数据挖掘
个性化模型
困惑度
Big data mining
Personalized model
Degree of confusion
