With the prevalence of the Internet and new media channels, consumer reviews have become one of the main determinants of Consumers’ purchasing decisions. This paper uses the Latent Dirichlet Allocation (LDA) model to identify the key factors that are of major concern to consumers, including design factors, laptop setup factors, logistics factors, after-sales factors, and user experience factors. And, we classify these factors into product quality factors and supporting service factors for new products. We then explore the relationship between online reviews and purchase decisions under these different factors, and also further explore the impact of interactions between online review metrics on purchase decisions. Our findings suggest that the impact of online reviews on consumer purchase decisions also varies considerably across different consumer focus factors. In addition, we find that the impact of the interaction between online review features is complex. In particular, consumers do not follow the positive guidance and make purchase decisions as we would expect when confronted with a large number of positive emotional polarity online reviews. Meanwhile, the interaction between negative emotional polarity and variance of online reviews had no significant effect on consumer purchase decisions. The variance of online reviews has a limited role in reducing consumer risk perceptions triggered by negative emotional polarity. Our study provides new evidence for the study of the impact of online reviews through text mining.

In the context of subsidy phase-out and the tightening of credit trading policy tools, enterprises producing internal combustion engine passenger cars have also started to produce new energy passenger cars, creating a competitive market with new energy passenger car enterprises. To explore the operation strategies and profit fluctuations of two enterprises in passenger car market and credit trading market, this paper develops a duopoly competition game model. Based on solving a multi-objective equilibrium problem, and performing numerical simulations with real data on Chinese passenger car market, this paper analyzes the interactive impact of policy tools on market competition. The conclusions point out that the numerical relationship between the price of positive new energy credits and credit deficit penalty is decisive to enterprises' pricing strategies. As the subsidy recedes, the competitiveness of new energy passenger car enterprise will decline. The policy tools at corporate average fuel consumption credit regulation level only have a negative interactive effect on hybrid energy passenger car enterprise's pricing and production strategy. There is a threshold for the difference between corporate average fuel consumption up-to-standard value and actual value, below which hybrid energy passenger car enterprise will produce more new energy passenger cars than internal combustion engine passenger car. The tightening of policy tools at new energy credit regulation level will drive both enterprises to raise price and reduce yield.