Natural Language Processing

The Natural Language Processing area studies how to build intelligent systems that can understand, generate, and act on human language. Research in this area treats language as a core component of broader intelligent systems, enabling artificial agents to interpret their surroundings, communicate with humans, and operate in complex physical or digital environments. Research emphasizes efficient training and deployment of large machine learning models; the development of specialized hardware to accelerate these models; integrated reasoning across multiple modalities; and decision-making mechanisms that enable systems to access and reason about external information or tools. By bringing together algorithms, systems, and application contexts, the area's focus on language-enabled intelligence reflects the demand of contemporary AI systems such as large language models and interactive agents.

Active research topics in this area include simultaneous translation, sign language translation, natural language-based navigation in 3D environments, vision-language multimodal models, universal semantic metalanguage, and reasoning and explanations by language models.