Parallel collaborative filtering recommendation model based on expand-vector

The recommendation system based on collaborative filtering is one of the most popular recommendation mechanism. However, with the continuous expansion of the system, several problems that traditional collaborative filtering recommendation algorithm (CF) faced such as speedup, and scalability are worsen. In order to address these issues, a parallel collaborative filtering recommendation model based on expand-vector (PCF-EV) is proposed. Firstly, the eigenvector is expanded reasonably to get the expand-vector based on the expand-vector model. Then, based on the expand-vectors, a series of similarity calculations are expressed. Finally the nearest neighbor item is found and a more accurate recommendation to the target user is given based on the calculation results. On the basis of these, the further optimization makes it applied to the parallel computing framework successfully. Using the MovieLens dataset, the performance of PCF-EV is compared with that of others from both sides of recommendation precision and the speedup ratio. Through experimental results, which are compared with CF, PCF-EV overcomes the problem of cold startup which the CF encounters. Moreover, the accuracy and recall ratio has been doubled. Compared with the serial implementation on the high-end dual-core CPU, the parallel implementation on the low and middle-end GPU reaches nearly 170 times speedup in optimal conditions.