Trip Recommendation Algorithm Based on Attraction Tags
Many route recommendation algorithms have been presented recently,its necessary to consider users preference in the recommendation.A travel recommendation algorithm is proposed based on visitor preferences.It analyzes the users preference for different types of attractions and forms a user-preference matrix.Then it calculates an initial clustering center based on the interest distributed by K-means algorithm,and establishes a neighboring set for the target user to score the historical users route value for target user by target users reference on scenic spot type distribution.The method finds the historical user route with the largest value for the target user,thereby generating the trip recommendation.The experimental results show that the algorithm can quickly calculate the smaller neighboring users and obtain the recommended results.It not only has faster recommendation efficiency,but also has better recommendation accuracy.It provides a good service on personalized route recommendation.
personalized recommendation user interest k-means algorithm
Dupeng Wang Biyang Ma Langcai Cao
Dept.of Automation,Xiamen University,Xiamen 361102,P.R.China
国际会议
西安
英文
361-367
2018-09-21（万方平台首次上网日期，不代表论文的发表时间）