Data analytics, decision support and upholding business ethics in Chinese automobile market

Ding, Ruixin (2025) Data analytics, decision support and upholding business ethics in Chinese automobile market. PhD thesis, University of Glasgow.

Abstract

After more than 70 years of development, Chinese automobile market has gradually become the world’s largest automobile market. Especially after 2010, the rapid development of new energy vehicles has significantly changed the market competition pattern. Under the impact of new brands and new car models, all automakers have to rethink the applicability of the original business theory in the new environment in order to maintain the original market advantage or obtain new market competitiveness. At the same time, with the prevalence of social media, online reviews have gradually become an important source of information for consumers to make decisions and for enterprises to analyse consumer demands, but it has also gradually derived the unethical business behaviours of manipulating fake reviews. In addition, the rise and application of artificial intelligence technologies such as machine learning and deep learning have proved their important potential in business analytics. This research aims to address the above business challenges by applying suitable machine learning methods, and to provide theoretical support and methodological practice for business decision-making and ethical maintenance. The overall research consists of three subdivision studies. In the first study, it is mainly discussed what important business data are available in the automobile market and their potential applications, and finally a comprehensive automotive industry dataset is created. The second study investigates the impacts and changes of first-mover advantages and country-of-origin effects in the Chinese automotive market through a novel sales analysis framework. It demonstrates the positive impact of the first-mover advantage and country-of-origin effects on sales performance, as well as that the first-mover advantage can be broken by innovative latecomers. In the third study, a large number of fake reviews were identified by the detection method based on BERT and PU-Learning we proposed, thus completing the study on the timing of manipulation of positive fake reviews. It shows the correlation between the timing of manipulation and car sales, brand sales, market size, and the duration of brands and models in the market. The whole research not only extends theories related to the first-mover advantage, country-of-origin effects, and the timing of manipulating fake reviews, but also demonstrates the superior performance of machine learning methods in business analytics. In addition to providing decision-making support for automakers’ corporate strategies and marketing strategies, this research has important implications in formulating industrial policies, protecting consumer interests, enhancing the credibility of online review platforms, and maintaining a fair business environment.

• Item Type: Thesis (PhD)
• Qualification Level: Doctoral
• Subjects: H Social Sciences > HD Industries. Land use. Labor > HD28 Management. Industrial Management; H Social Sciences > HF Commerce
• Colleges/Schools: College of Social Sciences > Adam Smith Business School
• Supervisor's Name: Chen, Dr. Bowei and Wilson, Dr. James
• Date of Award: 2025
• Depositing User: Theses Team
• Unique ID: glathesis:2025-85163
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 06 Jun 2025 13:13
• Last Modified: 06 Jun 2025 13:15
• Thesis DOI: 10.5525/gla.thesis.85163
• URI: https://theses.gla.ac.uk/id/eprint/85163

Related URLs

• Conference proceeding