Collaborative Filtering Recommendation Algorithm Based on label of tourist spots and User Preference

The traditional collaborative filtering recommendation algorithm relies on the user's scoring relation to the item. However, user's behavior data in the filed of tourism industry is sparsely, and the interaction between the data are few. These problems lead to the traditional algorithms are lacking of ability to acquire the users' preference, and influence the recommendation quality of tourist spots. In this paper, a collaborative filtering recommendation algorithm based on tourist spots labels and user preferences were proposed. By using the label information of the tourist spots, to extract visitors 'interest factors of spots and preferences weights, used the adaptive algorithm of neural network to optimize the users' preference feature vector, and computed the similarity between users to get the recommended result. The results show that compared with the traditional recommendation method, the method of this paper can ameliorate the accuracy of user similarity relationship and has a great improvement in the recommendation quality of tourist spots.