An Overview of Recommendation System Based on Generative Adversarial Networks
Biao Du, Lin Tang, Lin Liu · 2020
With the growth of explosive information resources, information overload has become increasingly serious. Users are eager to find what they need in a large amount of information resources. The recommendation system is an effective solution for filtering information and alleviating information overload. Traditional recommendation algorithms have achieved great results in applications and are widely used, but they also face serious data sparse problems. The proposal of the generative adversarial network provides another idea for modeling implicit feedback information-adversarial learning. By reviewing the research of recommendation algorithms based on generative adversarial networks in recent years, we analyze the differences and advantages of generative adversarial networks and traditional recommendation algorithms, and summarize the practical applications of these algorithms. Finally, the future of recommendation algorithms is discussed based on generative adversarial networks prospects for development trends.