Tourist Routs Recommendation Based on Latent Dirichlet Allocation Model

Abstract

Tourism is an indispensable part of our life nowadays. At the same time, DIY tours become more and more popular. Traditionally, people have to spend a lot of time browsing websites and reading travel notes to select a suitable tourist route. With the help of tourist routes recommendation system, people can obtain their tourist routes satisfying their demands automatically. We improve a tourist routes recommendation system which based on Latent Dirichlet Allocation (LDA) model. The recommendation system firstly uses LDA model to dig out the hidden theme from a large number of documents. Then, by using Collaborative Filtering algorithm, grades are generated for each user to each travel routes. In this way, we can determine which route is most suitable to the user clearly. Our evaluation results indicate that our recommendation system is effective and has high level of satisfaction with user's hobbies and interests.