A Survey of Multi-Label, Multi-Class Text Classification Techniques in Healthcare Recommendation Systems

Multi-label, multi-class text classification (MLMC-TC) is crucial for healthcare recommendation systems, enabling automatic extraction of complex clinical concepts from textual data. This survey presents a comprehensive review of MLMC-TC techniques applied to healthcare, covering traditional machine learning, deep learning, and hybrid approaches. Challenges such as label imbalance, noisy annotations, and evolving clinical taxonomies are discussed. Key datasets, evaluation metrics, and healthcare applications like clinical coding and specialist recommendation are highlighted. Finally, future research directions focusing on scalability, interpretability, and privacy-preserving methods are proposed.