This document presents a machine learning approach to detecting and preventing ulcerative colitis (UC) using colonoscopy videos and patient data. The authors extract features from colonoscopy footage and medical records, then train an ensemble learning model using decision trees, K-nearest neighbors, naive Bayes, and support vector machines. The model analyzes normal values and predicts UC levels. It aims to help diagnose UC and increase patient awareness. Evaluation shows decision trees performed well for text data while support vector machines worked best for image data. The framework could help tackle UC given its impact on many people worldwide.