新能源汽车普通消费者参与的大数据研究——基于文本挖掘和深度学习

Big Data Research on the Participation of Ordinary Consumers in New Energy Vehicles: On the Basis of Text Mining and Deep Learning

以汽车之家网站上新能源汽车的口碑评论文本为数据,基于文本挖掘和深度学习对国内新能源汽车行业的普通消费者参与进行了大数据研究。基于词云主题演化分析发现:国内普通消费者较为重视一些表层的因素,例如:“外观”“内饰”“空间”“性价比”等。但是对深层的因素重视不够,例如“电池(能耗、电耗、油耗)”“电机”“电控”“动力”等。另外,根据情感演化分析发现:“外观”“内饰”“空间”等维度的积极情感占比随着时间呈现上升态势,而“能耗”维度的积极情感倾向占比呈现下降态势,“动力”维度的积极情感倾向占比呈现先降后升态势。最后,针对这些大数据分析结果,基于价值共创理论为国内新能源车企提出了管理建议。

Based on text mining and deep learning, this paper conducts a big data study of the participation of ordinary consumers in the domestic new energy vehicle industry with the word-of-mouth review texts of the new energy vehicles on Autohome website. Through the analysis of the theme evolution of word cloud, it is found that ordinary consumers pay more attention to some surface factors, such as "appearance", "interior", "space", "economy" and so on. In contrast, undue attention is paid to the deep factors, such as "battery(energy consumption, power consumption, and fuel consumption)", "motor", "electric control", "power" and so on. According to the emotional evolution analysis, it can be found that the proportion of positive emotional tendency for "appearance", "interior", "space" and other dimensions increases with time, while that for "energy consumption" decreases, and that in the "power" dimension shows the first decline and following rise trend. Finally, some management suggestions are put forward for the new energy vehicle enterprises from the value co-creation theory.