An Analysis Framework to Reveal Automobile Users’ Preferences from Online User-Generated Content

Abstract

This work attempts to develop a novel framework to reveal the preferences of Chinese car users from online user-generated content (UGC) and guides automotive companies to allocate resources reasonably for sustainable design and improve existing product or service attributes. Specifically, a novel unsupervised word-boundary-identified algorithm for the Chinese language is used to extract domain professional feature words, and a set of sentiment scoring rules is constructed. By matching feature-sentiment word pairs, we calculate car users’ satisfaction with different attributes based on the rules and weigh the importance of attributes using the TF-IDF method, thus constructing an importance-satisfaction gap analysis (ISGA) model. Finally, a case study is used to realize the framework evaluation and analysis of the twenty top-mentioned attributes of a small-sized sedan, and the dynamic ISGA-time model is constructed to analyze the changing trend of the importance of user demand and satisfaction. The results show the priority of resource allocation/adjustment. Fuel consumption and driving experience urgently need resource input and management.