Dataset Improving Fast-Fashion Mobile Application Strategy Through Machine Learning-Based Sentiment Analysis

This dataset was constructed by scraping user reviews from the Google Play Store (applications: Zara, H&M and Uniqlo) using the google-play-scraper Python library, capturing extensive metadata before narrowing the focus to essential fields: review text, rating and sentiment label. After filtering out unnecessary columns, each review was manually annotated into one of three categories: positive, neutral or negative. The final dataset includes only three key columns which is the Review_Text, Rating, and Sentiment and is later designed for training and evaluating traditional machine learning classifiers: Naïve Bayes, K‑Nearest Neighbors (K‑NN) and Support Vector Machine (SVM). This dataset enables benchmarking of sentiment analysis performance on fast‑fashion mobile app reviews with specific applicability to strategy improvement, UX refinement and product feature prioritization based on user feedback.