Research on content analysis and quality evaluation for online reviews of new energy vehicles

Abstract

The development of China's new energy vehicle industry has been gradually transformed from past policy support to the survival of enterprises determined by the market, which forces the new energy vehicle industry to accelerate towards the market-oriented direction. We collected review data from online automotive forums and generated a corpus after pre-processing. Next, we extracted the statistical and topic semantic features of online reviews for new energy vehicles and clarified the quality evaluation indicators for online reviews. Then, the entropy weight-TOPSIS method was used to determine the indicators’ weights and assign values to the quality of online reviews. Finally, the quality scoring rules of online reviews were sorted out from multidimensional perspectives. The research results have shown that the statistical features and topic semantic features of online reviews on new energy vehicles can be used as evaluation indicators to construct a quality evaluation system. It was also found that the quality scores of lower-priced vehicle reviews tend to be consistent in different price groups, while the quality scores of mid-priced vehicle reviews are the most dispersed. The quality of online reviews varies significantly among other groups or models. The study results provide a simple criterion for consumers’ cognition of the value of online reviews. The results will enhance the ability of new energy vehicle companies to carry out online marketing with the help of online reviews and transform the purchase intention of potential consumers into actual purchase behavior.