Online Volume Exploration Based on Latent Dirichlet Allocation Model: A Case Study of Xiaomi and Huawei

Abstract

This study employs a multi-platform data analysis approach to compare the online public opinion landscapes of Huawei and Xiaomi. By gathering and examining data from major Chinese social media and e-commerce platforms, the research utilizes sentiment analysis and topic modeling methodologies. The findings indicate a marked divergence: Huawei generates significantly higher online volume and more positive user sentiment, largely driven by its focus on technological innovations such as imaging systems and chip technology. In contrast, Xiaomi’s presence, while substantial, is more defined by value-for-money propositions and design appeal. Users' discussions mostly revolve around price advantages and appearance experience. Furthermore, through user profile analysis, the study identifies that Huawei users are more inclined to be a group of tech enthusiasts with higher education and income, whose interaction mode is mainly focused on technical discussions. In contrast, Xiaomi users are mostly young consumer groups, and their interactions tend to be more about sharing usage experiences and price comparisons. The study also identifies distinct user demographics and engagement patterns for each brand. These insights reveal fundamentally different public opinion driving models and offer valuable empirical evidence for technology companies aiming to optimize brand communication and manage digital reputation in competitive markets.