Code Switched

Code-switching, the intermixing of two or more languages within a single utterance, is a prevalent linguistic phenomenon increasingly studied using computational methods. Current research focuses on developing robust machine learning models, often employing transformer architectures and techniques like hierarchical feature fusion, to address tasks such as automatic speech recognition, machine translation, and sentiment analysis in code-switched data. These efforts are driven by the need for improved natural language processing tools that accurately handle the complexities of multilingual speech and text, impacting fields like healthcare (e.g., autism detection), political discourse analysis, and cross-lingual communication. Overcoming data scarcity through techniques like data augmentation and synthetic data generation is a major challenge and focus of current research.