Cross-Platform Social Media and Agent Performance Dataset of Baylink (2020-2024)

Cross-Platform Social Media and Agent Performance Dataset of Baylink (2020-2024) Description This dataset comprises 5,435 samples of unstructured textual data collected from four major platforms—Instagram, Facebook, TikTok, and the Google Play Store—spanning the period from January 1, 2020 to December 31, 2024. The social media and app store data were acquired using automated tools: the Apify API for TikTok, and the google-play-scraper Python module for Google Play. For Instagram and Facebook, public comment data was harvested within ethical scraping boundaries. Each data point includes: Comment text (user-generated content) Timestamp (date of posting) These raw inputs were aggregated and cleaned using Python scripts, with the primary focus on preserving only the relevant unstructured comment text and its associated date metadata. No personally identifiable information (PII) was retained during collection or processing. In addition to social media data, the dataset includes annual unstructured records of Baylink agent activities from Indonesian Bank Meratera’s official reports for the same period (2020-2024). These reports were parsed and filtered to retain relevant qualitative insights for further natural language processing or organizational performance analysis. This dataset might be beneficial for: Sentiment analysis Temporal trend analysis Social media mining Customer feedback modeling Comparative behavioral studies between platforms Institution-specific service monitoring and evaluation (e.g., agent/user feedback modeling) Sources Instagram Facebook TikTok (via Apify API) Google Play Store (via google-play-scraper) Bank's Official Annual Reports (2020–2024) Temporal Coverage Start date: January 1, 2020 End date: December 31, 2024 Format .json, .csv, or .txt (you can specify the actual file type) UTF-8 encoding Two fields: comment_text, date Methodology All sources were scraped or extracted using automated Python pipelines, and unified in a post-processing stage. The unstructured agent data was manually digitized and semi-automatically parsed using regular expressions and NLP utilities. License and Usage Restrictions This dataset is provided strictly for research and academic purposes. Commercial use or any monetization of the dataset is strictly prohibited. Redistribution of modified versions must include appropriate attribution to the original source and must maintain the same license and usage terms. Keywords Social media, unstructured data, Apify, TikTok, Google Play, comment analysis, customer feedback, natural language processing, text mining, agent performance