The Design And Implementation Of Movie Recommendation System Based On Spark

The rapid development of internet technology had brought an age of information explosion, and the amount of information we faced grew in an exponential manner, which led to the information overload. On the one hand, information overload made it difficult for users to find what they really liked. On the other hand, lots of network information became "dark information",which can't be found by anyone. At present, the best solution for information overload was recommendation system. Recommendation systems generated recommendations for specific user by building the binary relation between user and item and finding the potential preference with the history. However, each recommendation algorithm had its weakness. Hence, hybrid recommendation system became a hot research area.This paper analyzed the common recommendation techniques at first and summarized their strengths and weaknesses. Based on different hybrid methods, two hybrid recommendation system designs were proposed in this paper, switching and feature augmentation, to explore the effect for recommendation performance. Experiments showed that hybrid methods could improve the recommend performance to some extent. In order to improve the ability of processing mass data, this paper designed and implemented movie recommendation system with feature augmentation hybrid recommendation algorithm based on Spark platform. At last, I evaluated this system with dataset.