Research And Application Of An Online Video Website's Movie Recommender Algorithm

With the rapid development of information technology and Internet industry,the scale of information resources is increasing very fast.The cost of finding valuable information is increasing day by day,and the problem of information overload is becoming more and more serious.In this background,the recommendation system,as an efficient information distribution model,is becoming more and more important.At present,the research of the recommender system mainly includes two aspects: the research of the system and the research of the algorithm,and the main content is still the algorithm research.Recommender system algorithm research is an important part in machine learning and artificial intelligence.In recent years,with the rapid development of machine learning,the field of recommender system algorithm has also made great progress.It has become a hot topic of common concern in academia and industry.Research and study on the current recommender system algorithms which are widely used,arranges common recommender system architecture and evaluating indicator.And point out recommender system algorithm which is based on similarity principle only concern the similarity relation between users or examples,while ignoring the information of user behavior sequences.Aiming at this problem,propose a new IA LSTM(Input Attention LSTM),which can extract sequential information,and takes online video website movie recommendation as an application scenario to conduct online effect testing.IA LSTM algorithm is based on Long Short-Term Memory recurrent neural network(LSTM),depends on the structural characteristics of recurrent neural network itself,it is good at extracting sequence information.Use Word Embedding technology to obtain the implied ideological features of film case and similarity information,and add an attention mechanism in the algorithm input,it has improved optimization of the network structure,so as to obtain better recommendation effect.Finally,compares the most widely used recommender system algorithm based on collaborative filtering and IA LSTM algorithm in the most commonly used datasets,such as Netflix dataset,Movie Lens100 M dataset and online application scenario of a video website.Analyze the comprehensive data,IA LSTM algorithm is better than collaborative filtering recommendation algorithm,whether it is recommended accuracy in common datasets or in the actual video website.It proves the effectiveness of IA LSTM algorithm.