Research On Product Competitor Identification And Sales Forecast Based On Consumer UGC

The development of the Internet has led to a large number of product transactions on the Internet,which has also spawned a number of companies focused on the Internet.With the help of the Internet,companies display products online to consumers,consumers buy products on the Internet,and then consumers display the product usage experience on the Internet in the form of consumer reviews.Consumer reviews can not only help follow-up consumers make product purchase decisions,but also companies can get consumers' evaluations of products from online reviews of consumers,so that products can be improved in the future.Companies can also use product reviews to forecast product sales.In the product forecasting process,product sales are not only affected by historical sales,comment content,and network popularity.Due to the existence of competition,product sales are more affected by competitors' sales.This paper collects product review data from various Internet platforms such as shopping websites,social media and product communities.By means of word frequency co-occurrence technology,the co-occurrence matrix of target enterprise products and competitor products is established from massive data,and social network technology is used for visual analysis.From which the target company's specific competitive products are obtained,Then,the sales volume of the target product,the sales volume of the competitive product,the emotional value of the product,the heat and price of the product network,and the regression model of the product sales are established.Moreover,because the existing researches believe that the relationship between the factors affecting product sales and sales volume is more nonlinear,this paper also establishes a neural network model for predicting product sales by using the Tensor Flow neural network framework developed by Google.Through the comparative analysis of the established two models,it is found that the prediction error of the regression model is much smaller than the prediction error of the neural network model,indicating that the relationship between the factors affecting product sales and sales volume is linear,not nonlinear.At the same time,in order to verify the validity of the analysis of the factors affecting the sales volume and the rationality of the established model,the model is further analyzed and verified by using the real sales data of the product,which proves that the analysis of this paper is reasonable and effective.