Forum Discussion Categorization

This project aims to automatically categorize forum discussion texts into five predefined categories using natural language processing and deep learning techniques. Leveraging models like Gated Recurrent Units (GRU) and Bidirectional Encoder Representations from Transformers (BERT), we built a text classification system capable of understanding contextual semantics and sequence patterns in user-generated content. The workflow involved data cleaning, tokenization, embedding generation, model training, and evaluation using accuracy and F1-Score. The model achieved high classification performance and ranked 1st in a university-hosted Kaggle competition, demonstrating its effectiveness in handling real-world, multi-class text classification tasks.