Abstract:

In recent years, the Shine-Muscat grape has rapidly expanded in the Chinese market due to its unique taste. However, the expansion of its cultivation scale has led to issues such as quality differentiation and price fluctuations, which have adversely affected the healthy and sustainable development of the industry. This study systematically explores public sentiment dynamics regarding Shine-Muscat grapes based on a large dataset of Weibo posts, using Latent Dirichlet Allocation (LDA) for topic modeling of comment texts and SnowNLP for consumer sentiment analysis. The findings reveal that public attention is primarily focused on price declines, taste variations, and pesticide residue disputes. Topic analysis identifies three core themes: variety characteristics and inter-variety comparisons, sensory quality and safety, and price and consumption experience. Sentiment analysis indicates that negative sentiments slightly outnumber positive ones, with negative emotions primarily stemming from dissatisfaction with taste, declining quality, and safety concerns, while positive evaluations highlight characteristics such as seedlessness and juiciness. By understanding public concerns and sentiment tendencies, this study provides a reference for the rational planning of the development of the Shine-Muscat grape industry, the adjustment of cultivation practices, and the improvement of product characteristics.

Key words: Shine-Muscat grapes, public opinion mining, topic identification, sentiment analysis