Collaborative filtering model of book recommendation system

With the rapid development of information technology and internet, people from an era of scarcity gradually entered the era of information overload. For information-consumers, finding themselves interested in information from a large amount of information is a very difficult task. As regard to information producers, letting the production information stand out and getting the attention of the masses of users is also a very difficult task. To solve this contradiction, first, we establish a decorrelation principal component analysis model based on the correlation theory to obtain the main factors affecting the user evaluation of books. Secondly, we establish a predictive scoring system based on linear regression theory which can predict book ratings. Finally, we establish a collaborative filtering model of book recommendation.