基于LDA模型的二手交易电商平台APP用户需求挖掘

User Demand Mining for Second-Hand Trading E-Commerce Platform APP Based on LDA Model

摘要: 随着共享经济与循环消费模式的兴起，二手交易电商平台迅速普及，但其发展仍面临用户需求复杂、服务同质化等挑战。为精准识别用户核心诉求，本研究基于LDA主题模型对二手交易平台APP的用户评论进行挖掘与分析。通过对11,299条有效评论进行文本清洗、分词与主题建模，识别出用户关注的三大需求主题：质检服务的专业化与透明化、价格评估机制的公平性与合理性、售后支持的可靠性与响应性。进一步结合SnowNLP情感分析发现，用户对售后服务的满意度最低(0.1213)，质检服务次之(0.3023)，整体情感倾向偏负面(0.2498)。研究表明，平台需在质检透明度、定价算法优化及售后体系重构等方面重点加强，以提升用户信任与体验。本研究为二手交易平台的精细化运营与功能优化提供了数据支持与理论参考。

Abstract: With the rise of the sharing economy and circular consumption models, second-hand trading e-commerce platforms have rapidly gained popularity. However, their development still faces challenges such as complex user demands and service homogenisation. To precisely identify core user demands, this study employs the LDA topic model to mine and analyse user reviews on second-hand trading platform applications. Through text cleaning, word segmentation, and topic modelling of 11,299 valid reviews, three primary demand themes emerge: the professionalism and transparency of quality inspection services; the fairness and reasonableness of price evaluation mechanisms; and the reliability and responsiveness of after-sales support. Further analysis using SnowNLP sentiment analysis reveals the lowest user satisfaction with after-sales service (0.1213), followed by quality inspection services (0.3023), with overall sentiment leaning negative (0.2498). Findings indicate that platforms must prioritise enhancing transparency in quality inspections, optimise pricing algorithms, and restructure after-sales systems to bolster user trust and experience. This research provides data-driven insights and theoretical references for the refined operation and functional optimisation of second-hand trading platforms.