Construction of the HMM Intelligent Recommendation Model and Its Application in the Film Ticketing System

A new intelligent recommendation method based on HMM is proposed in this paper, which can predict the future behavior of other users by training the known historical behavior data set of some users, so as to achieve the purpose of the intelligent recommendation. Firstly, the basic framework of an Intelligent Recommendation Model is established from the following aspects: the Representation of Observation Sequence, K-Means Clustering, HMM Training and HMM Nearest Neighbor Filtering. After that, the new model is applied to the Movie Ticketing System, which focuses on the Observation Sequence Representation of intelligent recommendation information in the HMM. Finally, basing on the MovieLens Data Set, the experiments of HMM Training and HMM Intelligent Recommendation in the Intelligent Recommendation Model are carried out. In the HMM Training experiment, the operation speed of the HMM Training is faster, and the HMM Models obtained by the HMM training is better. The results of the HMM Intelligent Recommendation experiments showed that the accuracy of the intelligent recommendation is 90.2%, meanwhile, the intelligent recommendation’s function of recommending movies for users and finding users for movies is realized. Therefore, the new Intelligent Recommendation Method based on HMM is effective and feasible, and this method is worth popularizing.