Natural Language Processing

Natural Language Processing (NLP) is a multidisciplinary research field within artificial intelligence (AI) that integrates principles from computer science, linguistics, cognitive science, and related disciplines. It aims to enable machines to understand, interpret, and generate human language across various societal domains, including education, healthcare, scientific research, finance, and beyond. 

At the IT:U NLP Lab, our overarching goals are twofold: (a) to develop robust and trustworthy language processing techniques informed by linguistic principles for understanding and generating human language; and (b) to build human-centered NLP applications that advance education, social sciences, humanities, and scientific discovery.

In pursuit of these research goals, our research agenda will center around the following topics:

• Governance of Large Language Models (LLM): we will focus on evaluating and mitigating undesirable LLM behaviors, such as building LLM evaluation benchmarks and understanding the behaviors of LLMs with real-world knowledge conflicts. 

• Computational Argumentation and Fact-checking: we aim to combine argument mining and fact-checking technologies to combat real-world misinformation, such as detecting health and science misinformation. 

• NLP for Science: the goal is to develop reliable cross-document NLP techniques to support scientific research activities, including generating slides from scientific papers, building scientific leaderboards, generating citation text, synthesizing biomedical studies, and analyzing causal relations between scientific research concepts.

• Knowledge and Reasoning: we aim to develop techniques that can effectively represent both explicit and implicit knowledge to perform complex reasoning tasks in situated contexts, such as designing a fallacious reasoning framework for misrepresented science and building an argumentation knowledge graph.

Members of this research focus have developed, or will develop, strong expertise in one or more of the aforementioned areas. We value demonstrable experience with open-source projects, and the ability to collaborate effectively with researchers from diverse interdisciplinary backgrounds. Prior experience with relevant areas of NLP and machine learning, along with strong engineering skills, is highly desirable. We offer a supportive environment for collaboration, including opportunities for international partnerships and industry engagement.