NCDs Listener: A Social Listening Tool for Non-Communicable Diseases

This dataset accompanies the NCDs Listener project, an open-source social listening tool designed to collect, process, and analyze public social media data related to non-communicable diseases (NCDs) The NCDs Listener leverages Natural Language Processing (NLP) techniques, the BERT language model, and Generative AI to extract insights from Thai and English content. Data is collected from publicly available posts using web scraping methods and API integrations, then processed through tokenization, stopword removal, normalization, lemmatization, keyword matching, and classification into thematic categories such as personal experiences, questions, and non-informative content. This dataset includes: 1) Raw and processed Facebook and Reddit comments 2) Model Development Data 3) Data from questionnaire responses of users The dataset is intended for research on public health communication, patient experience, and NCD awareness in online communities. It supports studies in medical sociology, health policy, and computational social science.