Abstract:

Based on data mining methods, traditional e-commerce recommendation algorithms work unreal-timely and imprecisely. In order to solve the problems, an improved algorithm ESER was presented, which utilized emotion energy sensory extensions to measure online-client emotion and to manage shopping tendencies. Then it classified ware characteristics and matched them with shopping information in a uniform emotion energy and ware characteristic space. Performance analysis and simulation results show that it can provide better real-time and customer satisfaction than the traditional does. And it can match customers and wares better than the traditional does.

Key words: distributed computing, emotion computing, e-commerce, recommendation algorithm, match