Abstract
Today many tourists read online tour reviews before planning their trips. However, it is stressful to thoroughly check online reviews because of a huge amount of information. In this paper, we propose a tour recommendation system based on online reviews. The system recommends tour reviews written by other tourists who have similar interests in tours to the user. The user’s interest in tours is extracted from a review and it is modelled as a document vector based on the Word2Vec model. Interest matching between the user and other travellers is performed based on the vector space model. The proposed system is tested on a data set including 259,084 tour reviews and compared with a conventional recommendation system which is based on the BoW (bag-of-words) model. The experimental results show that the proposed system is effective and promising.
