Sales Forecasting of Promotion Activities Based on the Cross-Industry Standard Process for Data Mining of E-commerce Promotional Information and Support Vector Regression

Sales Forecasting of Promotion Activities Based on the Cross-Industry Standard Process for Data Mining of E-commerce Promotional Information and Support Vector Regression

The purpose of the study is to analyze and forecast the effect of e-commerce promotional activities and sales volume, and improve the scientific and intelligent degree of enterprise management and decision-making through data mining and analysis. The short-term effect of e-commerce promotional activities was analyzed. In terms of effect of sales activities, the sales volume of goods during activities will be significantly increased. Meantime, as the activity time goes on, its ability to stimulate sales is gradually weakening, while the sales of goods before and after the activity will decline to a certain extent. The passenger flow will increase significantly during the activity period, while the passenger flow will not change significantly before and after the activity. The decrease of sales is mainly due to the low conversion rate during this period. The results show that the comprehensive prediction model has a certain improvement in prediction accuracy compared with the single support vector machine (SVM) prediction model. Combining with interest, effective association rules can be found better, and the decision-making model considering loss of profits before and after activities is helpful for enterprises to make better management decisions. Thus, under the guidance of current methodology, excellent prediction results and rich management inspiration can be acquired, which proves that the forecasting process proposed has great application value.