Research on personalized information recommendation platform for CSA users
Zhao Sheng1,Lu Yiping1,Qin Jing2
COMPUTER MODELLING & NEW TECHNOLOGIES 2014 18(12B) 174-183
1Hebei North University, Zhangjiakou,Hebei 075000, China
2HebeiUniversityofArchitecture,Zhangjiakou,Hebei 075000, China
This paper analyses the demand in personalized recommendation service and the characteristics of classification content for CSA users. Then it proposes a personalized information recommendation platform based on the interests of agricultural users. For the interest update in push service, we integrate mathematical modeling, three-dimensional synthetic techniques and the quantization techniques, to establish user interest model. For the distribution problem in push service of agricultural information, we propose an improved classification model based on genetic algorithm, BP neural network and multiple linear regressions. The process of feature extraction and algorithm implementation are also provided in this paper. The experiments show our recommendation algorithm based on user interest has obvious improvement in precision and recall compared to traditional algorithms. It can further excavate the users’ interest to cater to the preferences and make effective and in-time information recommendation.