It is a very significant issue for the e-commerce industry to recommend the high quality goods for the users in the mass information. However, the accuracy and efficiency have seriously affected the recommendation quality of the recommendation system. In order to solve this problem effectively, this paper first uses support vector machines (SVM) to classify products, obtaining the positive and negative feedbacks of each product. Then it calculates the comprehensive scores of product ratings and reviews for positive feedback; the comprehensive scoring data is used to establish a collaborative filtering recommendation algorithm, and a final recommendation list is generated according to the preference score. Finally, Taobao’s online data is used to verify the proposed algorithm. The research results show that the algorithm has good recommendation accuracy and speed, and has a certain practical value.
[1]
Shulin Liu,et al.
An Improved Collaborative Filtering Recommendation Algorithm
,
2016
.
[2]
J. Skilling,et al.
Algorithms and Applications
,
1985
.
[3]
Tian Da-gang.
Research on Chinese-text Automatic Classification Based on SVM
,
2006
.
[4]
Stan Matwin,et al.
Privacy-preserving support vector machine classification
,
2007,
Int. J. Intell. Inf. Database Syst..
[5]
Bin Wen,et al.
Collaborative Filtering Recommendation Algorithm based on Spark
,
2019,
International Journal of Performability Engineering.