Social Media Macroscope (SMM), A User-Friendly Open-Source Gateway for Social Media Research

Social Media Intelligence and Learning Environment (SMILE), a key application of the Social Media Macroscope (SMM) project, is an open-source platform tailored for social media research. SMILE addresses the limitations of existing social media analytics tools, such as high costs and restricted access to data acquisition and analysis methodologies. It offers a comprehensive suite of tools via an easy-to-use web interface, facilitating free academic research. By maintaining open-source code, SMILE ensures transparency and reproducibility of the research results. Additionally, it uses the user's credentials for data collection, adhering to social media platform protocols and guidelines to ensure proper data collection. One notable feature of SMILE is its sentiment analysis capability, which includes the VADER (Valence Aware Dictionary and sEntiment Reasoner) algorithm, SentiWordNet algorithm, and a machine learning-trained sentiment model with debiased word embeddings. This makes SMILE an invaluable tool for researchers aiming to understand and interpret the emotional tone of social media content. Other analytics offered by SMILE include natural language processing, network analysis, and machine learning classification, name entity recognition, and topic modeling, each incorporating multiple algorithms with appropriate academic citations. SMILE's microservice architecture ensures portability and scalability. CIlogon provides secure and lightweight identity management. The backend, powered by Node.js/Express and GraphQL, handles user requests and social media data collection. Analytics modules run in separate containers, with RabbitMQ managing communication between the backend and analytics modules. MinIO is employed to provide secure data storage and an integration with NCSA Clowder facilitates community data sharing. SMILE leverages Docker containers managed by Kubernetes and Helm charts for continuous integration and deployment. All images and deployment charts are accessible on GitHub and Docker Hub. Automation scripts, leveraging GitHub Actions, are in place to simplify build, publication, and deployment tasks, making migration, upgrade, and deployment processes efficient and user-friendly.