A Movie Recommendation Model Based On Paragraph2Vec

In recent years,Internet technology has been widely used,and the information on the network has accumulated rapidly,and the problem of information redundancy has become more and more serious.In order to solve such problems,a recommendation system has emerged.At present,the recommendation system has been applied to various fields such as music,film and television,commodities,and news.This article uses the movie review text to establish a movie recommendation model.First,the word vector and paragraph vector model in text mining technology are studied,and the paragraph vector model is used to map movie reviews to the vector space,which solves the calculation problem of similarity between movies;then focuses on the recommendation based on collaborative filtering.Algorithms and related theories and implementations of the Slope One recommendation algorithm.Experiments were performed to compare the effects of several recommendation algorithms.Among them,the improved Slope One recommendation effect is the best,which also proves that the similarity of the project is introduced into the traditional Slope One The algorithm has a certain improvement effect;finally,for the "tourist" users who will not leave a movie watching record,the paragraph2 Vec model is used to realize the movie recommendation function.