Abstract
Automobile purchase intentions of customers relate to car dealers’ costs and affect the car dealers’ marketing strategy and manufacturing process in the long term. Automobile purchase intention classification has become critically important for car dealers. In our paper, we innovatively constructed a hobby based network and a working based network of customers, and used customers’ profile of same group as inputs to the deep learning model to predict customers’ purchase intention based on community detection by social network analysis. Based on the real-world dataset, our experimental results verify that the framework with both hobby-based network and working-based network using deep learning method has best performance, which is 14% better than the baseline model. And the hobby-based network outperforms working-based network. Because of the advantage of consumer’s personality, hobbies can be used for better predicting the purchase intention. Therefore, our proposed framework is a potential tool for automobile purchase intention classification.
Recommended Citation
Yin, Han; Wang, Yue; Li, Qian; Xu, Wei; Yu, Ying; and Zhang, Tao, "A Network-enhanced Prediction Method for Automobile Purchase Classification using Deep Learning" (2018). PACIS 2018 Proceedings. 111.
https://aisel.aisnet.org/pacis2018/111