Trip Recommendation Algorithm Based on Attraction Tags

Many route recommendation algorithms have been presented recently, it' s necessary to consider user‘s preference in the recommendation. We propose a travel recommendation algorithm based on visitors preferences. It analyzes the user' s preference for different types of attractions and forms a user-preference matrix. Then It calculate 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 user‘s 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.