Incorporating LDA Based Text Mining Method to Explore New Energy Vehicles in China

Abstract

Tracking the evolution of policy and development of new energy vehicles (NEVs) in China is of critical significance, because it helps generate rational prediction regarding future trends. To this end, this paper investigated the 5185 articles on NEV obtained from China National Knowledge Infrastructure by means of latent Dirichlet allocation (LDA)-based text mining. Word count was performed to highlight important keywords for different periods of years of publication. In addition, topics were identified from the abstracts of these articles using LDA. Findings suggest that attention on NEV in China has been growing and will continue to grow in the predictable future. Full electric vehicle, being the currently dominating form of NEV, will continue to play the leading role. Meanwhile, China’s NEV industry requires further investment into charging, battery, personnel training, and patent portfolio.