A Prediction Study on E-commerce Sales Based on Structure Time Series Model and Web Search Data
With the development of e-commerce, online shopping has become a primary channel for consumers, and it is meaningful to predict the sales volume of e-commerce. This paper combines web search data and structure time series model to predict the womens clothing sales volume of Taobao. Firstly, explore the correlation of consumers web search behavior and purchase behavior theoretically; Secondly, eliminate the trend and the seasonal factors of sales volume using structure time series model and calculate the residual series. Then construct the search index and establish the prediction model based on search data and residual of sales volume. The result shows that the mean absolute percentage error of 7-days sales volume prediction is 4.84%.
search data structure time series sales prediction
Dai Wei Peng Geng Liu Ying Li Shuaipeng
University of Chinese Academy of Sciences, Beijing 100090
国际会议
长沙
英文
5346-5351
2014-05-31（万方平台首次上网日期，不代表论文的发表时间）