Automobile Sales Forecast Based on Web Search and Social Network Data
Yang Zipei, Chong Zhang · 2020
The information on the Internet will affect consumers’ purchase decisions, especially for high-value products such as automobiles, so some data from search engine and We Media platform to a certain extent will reflect the car sales trend. This paper applies the monthly sales data of Audi from 2011 to 2019 to explore the impact of Baidu and Weibo data on automobile sales forecast. By adding car sales data, Baidu search indexes of selected keywords, Weibo text amount indexes and Weibo sentiment indexes, three machine learning models are established to compare with traditional time series model, thus key indexes helping improve the prediction accuracy are analyzed and marketing advice will be given to car companies.