Abstract
Automotive forums offer abundant and genuine data from car enthusiasts, enabling larger-scale customer satisfaction research. Current studies utilizing forum data for customer satisfaction measurement tend to be coarse-grained due to the intricacy of automotive products. As such, we introduce an improved aspect attention-based BiLSTM with aspect embedding (AATAE_BiLSTM) model, precisely identifying users' aspect-level sentiment tendencies. Employing a hierarchical indicator system, we capture users' sentiment tendencies at three levels, from fine-grained to coarse-grained. Moreover, we present a novel comprehensive quantitative model of satisfaction and attention based on sentiment tendencies, facilitating the measurement of users' satisfaction and attention across different indicator levels. We analyze BYD's NEV word-of-mouth comments data from AutoHome to gauge user satisfaction with BYD NEVs. Our study offers valuable insights for NEV companies to enhance their products and increase competitiveness.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
