Efficiency Optimization Of Customer Service In Cybershop Based On Attribute Preferences And Genetic Algorithm

With the vigorous development of the Internet technology in recent years, the online shopping is accepted by more and more clients. The rise of e-commerce in China, especially the boom in entrepreneurial online, makes an online shop the first choice for the start-up. However, the popularity of online shops highlights the problem of the labor shortage. According to statistics, the number of online shop personnel gap is up to 5.321 million in 2015, in which the customer service staff takes up the biggest proportion. The increase in the number of online shops will render the gap bigger and bigger. The present situation makes it of great significance for all the online shops, even for the development of entire C2C E-commerce to figure out how to improve the efficiency of the customer service staff and reduce the impact in the industry brought by personnel gap.To fulfill the purpose of improving efficiency, this paper firstly introduces the importance of the research and the research status and explains the structure and content of the paper in a simple way. Next, in response to the present problems, this paper analyzes the daily work of the customer service staff through the study of social science methods and finds out that the key to improve the efficiency of the customer service staff is to change the laggard recommendation method used currently. After then the paper introduces the current condition of the personalized recommendation algorithm and some main personalized recommendation algorithms, analyzes the advantages and disadvantages of each algorithm to provide a reference for the personalized recommendation algorithm used in online shop. Based on the above analysis, this paper puts forward a new personalized recommendation algorithm on the basis of attribute preference and genetic algorithm for online shop. The new algorithm takes the content-based recommendation as algorithm guidelines, reflects the product utility value for customers by customer’s preference for product’s attributes and predicts the product utility by genetic algorithm to complete the recommendation. Finally, with the public data sets, we design the contrast experiment between the new algorithm and the recommendation algorithm based on contents. The prediction accuracy and the classification accuracy indicators of the contrast experiment shows that the new recommendation algorithm is superior to the content-based recommendation algorithm and can effectively improve the efficiency of recommendation.