Research On Automobile Sales Forecast Based On Online Word-of-mouth And Search Data

The automobile industry has a great influence on promoting employment and consumption,which to a large extent,promotes the development of Chinese economy,especially the manufacturing industry.At present,Chinese automobile market has entered a slow growth stage,even a negative growth stage.It is of great practical significance for Chinese automobile market to accurately predict the future sales volume of automobiles from both macro and micro perspectives.Sales forecasting is an essential element for implementing sustainable business strategies in the automotive industry.Accurate sales forecasts enhance the competitive edge of car manufacturers in the effort to optimize their production planning processes.The rapid development of the Internet has recorded the consumer's purchase behavior and left a lot of data.Under the background of Internet,how to use consumer behavior data to improve the prediction effect of car sales has become a key.A large body of research uses data from social media websites to predict offline economic outcomes such as sales.However,recent research also points out that such data may be subject to various limitations and biases that may hurt predictive accuracy.At the same time,a growing body of research shows that a new source of online information,search engine logs,has the potential to predict offline outcomes.We study the relationship between these two important data sources in the context of sales predictions.Focusing on the automotive industry,a classic example of a domain of high-involvement products,a benchmark model is constructed combining with economic variables,such as gasoline price and consumer confidence index.In addition to using Baidu search trend data,it also uses the automobile word-of-mouth data in the automobile home.A prediction technology is proposed to explore the interaction between search data and online word-of-mouth data in the prediction,and then predict car sales in a month.Finally,considering the different attributes of the car's word-of-mouth score,explore the attribute word-of-mouth that affects the prediction effect,in order to improve the accuracy of the prediction.In the sales prediction model based on machine learning algorithm,it is found that in high-involved products,search data provides digital traces left by consumers,which can be effectively used to enhance sales volume prediction.Enterprises can study theonline search behavior of consumers in almost zero cost way.Therefore,consumer Internet information search may become a valuable resource to understand the market demand.On the basis of the overall word-of-mouth data,it is also proved that the model with search data can increase the prediction effect of the total word-of-mouth data and significantly improve the prediction accuracy.This evidence shows that the sales related information embedded in search data is external and does not overlap with the corresponding information embedded in word-of-mouth.It is also found that adding the overall word-of-mouth score to the model can not effectively improve the prediction ability.However,the model considering the changes of word-of-mouth in various attribute features of the automobile can effectively improve the prediction effect of automobile sales,which provides a new research idea and method for mining effective prediction information from big data with the help of big data analysis method in the future.