Sentiment and attention of the Chinese public toward electric vehicles: A big data analytics approach

Abstract

Individuals' attention and sentiment are the keys to adopting electric vehicles (EVs). Traditional questionnaires and interviews cannot fully and accurately reflect the attention and sentiments. Social media interactions can provide a new data-driven perspective to explore the sentiment toward EVs. This study uses data from public posts on Weibo to investigate intersectionality in EV - sentiment and attention as per user, gender and region. On a 1,149,243-text corpus extracted from the Weibo posts, a computational social science methodology was employed with a mixed-method of deep learning and topic modeling through Latent Dirichlet Allocation algorithm. Results showed that attention toward EVs mainly comes from official users rather than individual users (IUs), and IUs' attention is closely linked with EV policy change. Additionally, the attention level and growth rate toward EVs vary across regions and men pay more attention to EVs. There exist significant differences in both positive and negative sentiment driving factors across genders. This study facilitates to EVs’ policy-making and strategy in China and other countries.