Affiliations: [a] College of Computer and Information, Hohai University, Nanjing, China
| [b] Faculty of Computer and Software, Huaiyin Institute of Technology, Huaian, China
Abstract: Conventional recommender systems of cold chain logistics distribution mainly focus on the recommendations of the source of cargos, refrigerator trucks and refrigerators in the supply and demand link of cold chain, but ignore contextual information such as time, position and user devices. In this paper, we analyze the contextual information on cold chain logistics distribution and propose a multidimensional context-aware recommendation algorithm(MCARA). MCARA firstly carries out fuzzy clustering on contextual information in historical data set and obtains the contextual clusters. In addition, MCARA compares current user context with historical contexts to get current contextual cluster, and selects out the data with same contextual clusters from historical data set. Finally, MCARA uses the user-based collaborative filtering algorithm to perform personalized recommendations. The simulation results show that MCARA can improve the forecast accuracy of cold chain logistics distribution, with about 10% improvement over other eight approaches.
