Year: 2025
Type: conference-paper
Abstract: Customer Satisfaction (CSAT) is a critical metric that measures customers' perceptions of products, services, and overall experience, and it directly impacts repurchase rates and brand image. This study analyzed 85,907 customer feedback records from real e-commerce platforms, selected and identified important features that affect customer satisfaction (such as handling time, agent shift, tenure bucket, category, etc.), and used regression analysis to explore the impact of these features on customer satisfaction and their significance, and analyzed the possible causes of the factors. In addition, in order to further help e-commerce platforms optimize customer service, this study also built a machine learning model to predict CSAT, and evaluated the importance and impact of these features through model interpretability tools.
Author: Haoyue Wu
Institution: South China Agricultural University
Language: English
Domain: Social Sciences
Open Access status: gold
OpenAlex ID: W4415074088