Predictive chatbot for student success: source code and database schema

This dataset contains the full source code, cron job scripts, and database schema used in the study “Early detection and personalized academic support using a predictive chatbot for student success” (submitted to PeerJ Computer Science). The system integrates a high-performance LightGBM predictive model with an AI-powered WhatsApp® chatbot based on the GPT-4o-mini model to identify at-risk students and deliver early, personalized academic support.

The repository includes:
• Node.js backend code for the chatbot, implemented using the Baileys WhatsApp Web API, Prisma ORM for PostgreSQL, and the OpenAI API (GPT-4o-mini) for conversational responses and semantic search over course syllabi.
• Automation scripts (cron jobs) that periodically trigger the prediction process, read the academic management database, and update the table of at-risk students used by the chatbot for proactive outreach.
• The complete SQL schema of the institutional database tables involved in the predictive pipeline and chatbot integration, including the structure of the enrollment, grades, and chatbot status tables.
• Configuration and documentation files (README) describing the required software versions, environment variables, and step-by-step instructions to deploy the system in a new institutional context.

For ethical and legal reasons, no real student records are included in this dataset. Instead, we provide the full database schema and example configuration files so that other researchers can adapt the code to their own institutional data while preserving student privacy. This dataset is intended to support transparency, reproducibility, and reuse of the proposed predictive-chatbot architecture in higher education settings.