Shi Yufan

School of Business, Xi'an International Studies University

Abstract：

With the deep penetration of e-commerce and social media, online user reviews have become an important text resource reflecting the preferences, attitudes and decision-making logic of high-end perfume consumers. As a luxury that relies on sensory experience and emotional value, perfume online consumption has a high degree of experience uncertainty, and consumers highly rely on existing comments to reduce purchase risk. This research takes 322 user comments of the JD platform "Miss Dior Blooming Bouquet Eau de Toilette" from June to November 2025 as samples, grabs data with octopus collector, and uses data cleaning, description statistics, word segmentation and word frequency statistics, TF-IDF keyword extraction, topic clustering, semantic co-occurrence, and emotion analysis methods to conduct text mining. The analysis gradually progresses from keyword recognition, topic induction, attribute association to emotional judgment. Research has found that the consumer evaluation system consists of three dimensions: purchase motivation, product attributes, and related services; The commenting behavior is significantly influenced by the promotion cycle and there are individual sensory preference differences. Emotion analysis shows that overall emotions are neutral to positive, with negative feedback concentrated on the duration and intensity of fragrance retention; Semantic co-occurrence further indicates that fragrance attributes are closely related to subjective emotional judgments and usage scenarios, collectively forming a complete perceptual evaluation context.

Key Words：

high-end perfume; online reviews; consumer reviews; Miss Dior Blooming Bouquet Eau de Toilette