基于Apriori改进算法的旅游个性化推荐  

Personalized travel search based on improved Apriori agorithm

作　　者：徐宏[1] 陈焰[1] 支艳利 张金鹏 

机构地区：[1]昆明理工大学信息工程与自动化学院,昆明650504 [2]云南财经大学信息管理中心,昆明650221

出　　处：《微型电脑应用》2018年第1期74-79,共6页Microcomputer Applications

摘　　要：针对推荐系统在旅游领域面临的旅游数据稀疏、复杂,用户难以准确表述自己需求,以及用户偏好信息难以获取等特殊问题,对传统旅游个性化推荐系统进行分析。并针对这些问题,分析当前Apriori算法,提出了一种结合矩阵聚类的Apriori算法(Matrix Cluster Apriori,简称MC-Apriori)。基于MC-Apriori算法提出了面向旅游用户个性化搜索的关键词推荐模型。通过对用户搜索和阅览的历史关键词频繁集挖掘,向用户推荐满足其当前搜索兴趣倾向的旅游信息。对收集的康辉旅行社的语料数据进行测试,并将MC-Apriori算法分别与原Apriori算法、传统的推荐方法比较,证明了MC-Apriori算法在保证推荐结果准确、多样和个性的前提下,还解决了旅游个性化推荐面临的特殊问题,提高了效率。This paper discussed the special problems faced by the personalized travel recommendation,which included the sparse and complex tourism data,the user expressed their needs difficultly,as well as the historical tourism knowledge and preferences of information was difficult to obtain.Faced to the difficult,MC-Apriori was given.A tourism-oriented key words recommendation algorithm in personalized was mining the frequent itemsets of the historical key words by MC-Apriori.The essential of proposed method was mining the frequent itemsets of the historical key words,and recommend to the user to meet the current search interest in the trend of the new key words.Finally,based on corpus of China Comfort Travel,compared with the traditional recommendation method,the MC-Apriori not only ensured the accuracy,diversity and individuality of the recommended results,but also solved the special problems of personalized travel recommendation,and improved the efficiency.

关 键 词：旅游推荐 MC-Apriori TextRank 关键词 搜索 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]