Research on personalized recommendation system of e-commerce intelligent customer service based on artificial intelligence technology

This paper proposes an intelligent recommendation method for electronic commerce. This paper aims to study the problem that the MAE value in the existing e-commerce recommendation algorithm is too large, which affects the prediction accuracy of the system. In terms of hardware, PC104 industrial computer is used as the processing CPU. DSP and memory chip MPC565 are used as processing center. The clock signal, reset circuit and power supply circuit are designed by means of signal fiber compensation. The software calculates the personalized characteristics of e-commerce according to its characteristics, and then uses the association rules algorithm to process the personalized characteristics data. Use Java tools to convert it into code, use different functions of the code to achieve intelligent recommendation and then complete the design of the software part. The experiment proves that the e-commerce intelligent recommendation model constructed by using individual characteristics has less sample set and higher prediction accuracy and practicability than the conventional e-commerce intelligent recommendation method.