Harvesting E-Commerce Data for Multimedia Recommendation Systems in the Gaming Industry

This repository contains the dataset and source code used in the experiments reported in the associated PeerJ publication. The repository contains a comprehensive dataset and supporting source code collected from the Google Play Store, focusing on game-related categories and keywords such as "action", "shooter", "battleground", and "learning games". The repository includes: • raw_data.json – Metadata for game applications, including app titles, ratings, installation counts, genres, and other relevant attributes. • rawReviews.json – User reviews and ratings collected for the corresponding applications. • Data Crawling Scripts – Python scripts used to scrape application metadata and user reviews from the Google Play Store. • Preprocessing and Cleaning Scripts – Python code used for data cleaning, filtering, preprocessing, and preparation for analysis. The dataset and source code are provided to support reproducible research in recommendation systems, multimedia recommendation, natural language processing, sentiment analysis, user behavior modeling, and machine learning applications involving mobile gaming platforms. Researchers may use these resources to replicate the experiments reported in the associated publication or to conduct further studies in related domains. Note: This repository supersedes the previous version by providing both the dataset and the complete source code for data collection, preprocessing, and cleaning in a single archive, thereby improving reproducibility, transparency, and ease of access for researchers.