Menu
Explore a research direction
Natural language processing
research labs.
Explore labs working on natural language processing, computational linguistics, and language models. Compare their listed interests and selected publications, then consult official pages for current projects.
What to compare
Check the languages, tasks, and evaluation methods in recent work. Ask whether the project focuses on models, datasets, linguistic analysis, or a particular application, and what preparation is expected.
Profiles appear here because their listed research keywords match this topic. Inclusion does not establish an open position. Check the official sources and any published application instructions before reaching out.
Prepare a research introduction ↗26 research-led lab profiles
Browse labs by their listed college affiliationAlan Ritter Research Group
Georgia Institute of TechnologyAlan Ritter · Associate Professor
How can machines learn to read vast quantities of text with minimal supervision? His research interests include natural language processing, information extraction, and machine learning. He was a postdoctoral fellow in the Machine Learning Department at Carnegie Mellon. His research aims to solve challenging technical problems that can help machines learn to read vast quantities of text. In a recent project, his group built a system that reads millions of tweets for mentions of new software vulnerabilities.
Arman Cohan Research Group
Yale UniversityArman Cohan · Assistant Professor of Computer Science
How can language models and representation learning be adapted to long-context, multi-document, and specialized-domain tasks while improving reliability and evaluation? The Yale NLP Lab studies language modeling, representation learning, retrieval, and domain-specific applications across mechanistic interpretability and evaluation. Researchers investigate capabilities and reasoning of large language models, retrieval-augmented methods, and knowledge-intensive environments using experiments and modeling. The group develops pretraining and summarization techniques for long and multi-document inputs and evaluates few-shot and faithfulness metrics for real-world NLP tasks. Lab members publish in major NLP and ML venues and maintain an active research and student training program.
Bing Liu Research Group
University of Illinois ChicagoBing Liu · Distinguished Professor
How can AI systems continue learning after deployment as the world and their conversations change? The group investigates lifelong and continual learning, autonomous AI and open-world learning. Research also spans natural-language processing, chatbots and data mining, with a longstanding focus on extracting opinions and sentiments from text. Documented work includes detecting deceptive opinions, learning from positive and unlabeled examples, and extracting information from the web. The current program examines self-evolving agents and continually learning dialogue systems, connecting language understanding with adaptation over time.
Chao Zhang Research Group
Georgia Institute of TechnologyChao Zhang · Associate Professor
How can machines understand text data in a more label-efficient and robust way in open-world settings? Chao Zhang works at the School of Computational Science and Engineering, Georgia Institute of Technology, and the research area is data mining, machine learning, and natural language processing. Specific research topics include weakly-supervised learning, out-of-distribution generalization, interpretable machine learning, and knowledge extraction and reasoning. The page lists research areas of knowledge representation and extraction, text mining, machine learning, and natural language processing. Affiliations on the page include the Institute for Data Engineering and Science (IDEaS) and Machine Learning (ML@GT).
Chenhao Tan Research Group
The University of ChicagoChenhao Tan · Associate Professor of Computer Science
How can AI explanations and interactions be designed so humans make better decisions with machine assistance? The Chicago Human + AI (CHAI) Lab builds algorithms to align AI explanations with human interpretation to improve human-AI decision making. The group analyzes large textual datasets and natural experiments to understand how language shapes human decisions, such as persuasion and bargaining. Researchers develop decision-focused summarization and delegation methods so AI highlights the most decision-relevant information for people. The lab explores few-shot learning from human explanations to enable people to efficiently improve large language models.
Chitta Baral Research Group
Arizona State University (Tempe)Chitta Baral · Professor
How can knowledge and reasoning be combined with statistical models to improve language and vision understanding? The laboratory of Cognition and Intelligence develops neuro-symbolic and knowledge-based approaches to natural language understanding, vision-language question answering, and multi-modal document understanding. Projects investigate prompting and programming large language models, bias origins in NLP, efficient low-data methods, and knowledge assimilation for inference. Researchers apply methods spanning symbolic knowledge representation, reasoning frameworks, and machine learning to develop robust, generalizable models for tasks in cybersecurity, robotics, and biological and health sciences. The group maintains publications and a public lab website documenting tools and project outputs.
Daniel Fried Research Group
Carnegie Mellon UniversityDaniel Fried · Assistant Professor
How can people use language to get computers to carry out useful tasks in interactive settings? Daniel Fried focuses on enabling people to use language to get computers to carry out useful tasks in grounded and interactive settings. One recurring theme in his work is pragmatics, which views language as an action that people take in context. Recently he has been focusing on code generation, aiming to make programming more communicative. He also works on LLM agents, aiming to make them better collaborators.
Daphne Ippolito Research Group
Carnegie Mellon UniversityDaphne Ippolito · Raj Reddy Assistant Professor
What are the tradeoffs and limitations of generating text with large language models? Daphne Ippolito is an Assistant Professor in the Language Technologies Institute at Carnegie Mellon University. Her research focuses on privacy and security issues arising in language generation systems. She also studies how training data affects what models can be trained and how to build better systems for realistic evaluation. She holds affiliate positions in the Machine Learning Department and the CyLab Security & Privacy Institute.
David Wagner Research Group
University of California, BerkeleyDavid Wagner · Professor
How can computer security keep up as artificial intelligence changes how systems are built and attacked? David Wagner is currently working on security for AI, particularly security for large language models, AI for security, and other topics in computer security. He has previously worked on software security, electronic voting, wireless security, sensor network security, and applied cryptography. He is part of Berkeley's security research group and co-director of the KACST-UCB Joint Center on Cybersecurity. Released software includes PrimeVul, a high-quality dataset of vulnerabilities in code, and DiverseVul, a large dataset of vulnerabilities in code.
Graham Neubig Research Group
Carnegie Mellon UniversityGraham Neubig · Associate Professor, Language Technologies Institute
How can computers understand and translate human language well enough to let every person communicate in their own language? Research addresses NLP and large language models, spanning foundational gains in what models can do and uses such as software development. One thread on core technology aims to improve how accurately systems parse, translate, produce, or answer textual inputs, notably through machine translation and language understanding. Another thread models language alongside other modalities, working with speech and with the connection between human languages and formal ones like programming code. A third thread builds models using naturally occurring data and unsupervised, semi-supervised, and active learning approaches, cutting reliance on costly manual annotation.
Humanitarian Informatics Lab
George Mason UniversityHemant Purohit · Associate Professor of Information Sciences and Technology
During a crisis, useful information can be buried in a flood of online messages. Hemant Purohit develops human-centered AI systems for analyzing behavior and managing unstructured, multi-source data in real time. His Humanitarian Informatics Lab combines natural-language processing, semantic computing, and machine learning with social and cognitive theories. Applications include emergency preparedness and response, public safety, and cybersecurity. The research emphasizes human control and collaboration with intelligent systems, offering an intersection of data science, public services, and crisis informatics.
Jiawei Han Research Group
University of Illinois at Urbana-ChampaignJiawei Han · Michael Aiken Chair Professor
How can large collections of data and text be mined for structure and knowledge? Jiawei Han is a Michael Aiken Chair Professor in the Siebel School of Computing and Data Science at the University of Illinois at Urbana-Champaign. His page lists research in data mining, text mining, and intelligent systems. His current research listings include an NSF EAGER project on multi-hop reasoning with LLMs using a structure-guided approach. He also teaches a course on text mining with large language models and has written the textbook Data Mining: Concepts and Techniques.
Lillian Lee Research Group
Cornell UniversityLillian Lee · Charles Roy Davis Professor
How can natural language processing contribute to the analysis and facilitation of socially embedded processes? Lillian Lee is most interested in connections between natural language processing and social interaction. Her areas of work in language-based interaction include identifying pivotal moments in mental-health therapy conversations. Her areas of work also include grammar formalisms and parsing, such as dependency parsing and related formalisms. Other areas include multimodality and LLMs, such as the New Yorker Caption Contest as a humor benchmark.
Maarten Sap Research Group
Carnegie Mellon UniversityMaarten Sap · Assistant Professor
How can NLP systems be given social intelligence and an understanding of social biases in language? Maarten Sap is an Assistant Professor at Carnegie Mellon University's Language Technologies Institute. His research centers on enhancing NLP systems with social intelligence and understanding social biases in language. He investigates narrative dynamics, analyzing how stories shape and reflect social contexts. He also studies the complexities of toxic language in order to improve AI moderation tools.
Mark Dredze Research Group
Johns Hopkins UniversityMark Dredze · John C. Malone Professor
How can natural language processing of social media and clinical text detect and monitor public health outcomes in real time? The group develops artificial intelligence systems that apply statistical models of language to social media analysis for public health surveillance. Researchers design and evaluate statistical methods for information extraction addressing syntax, semantics, sentiment, and spoken language processing tasks. They build new clinical natural language processing methods to analyze medical records for applications in clinical informatics and medicine. Applications studied include tobacco control, vaccinations, infectious disease surveillance, mental health, drug use, and gun violence prevention using AI-based language models.
Marti A. Hearst Research Group
University of California, BerkeleyMarti A. Hearst · Professor
What does research on information visualization, scholarly document reading and search, and computational linguistics involve? Marti A. Hearst describes her research as focused on information visualization, scholarly document reading and search, and computational linguistics. Her listed research areas also include search engines and their user interfaces and information-rich user interface design. She is the author of Search User Interfaces, which her page calls the first academic book on this topic. Her courses have included information visualization and presentation, applied natural language processing, and user interface design.
Mohit Iyyer Research Group
University of Maryland, College ParkMohit Iyyer · Associate Professor
How can large language models support long-form text generation and long-context language understanding? Mohit Iyyer is an associate professor in computer science at University of Maryland, College Park and a member of CLIP. His research focuses on natural language processing and large language models. Current topics of interest include AI agents for collaborative writing and AI-generated text detection. A recent demo release is IdeaLens, a model for detecting whether the ideas in a text came from a human or AI.
Percy Liang Research Group
Stanford UniversityPercy Liang · Thomas M. Siebel Professor of Machine Learning
How can language models become more open, accessible, and understandable to the researchers who study them? The group concentrates on foundation models, particularly language models, using open-source development and rigorous benchmarking to make them more accessible and understandable. Earlier work covered robustness, interpretability, human interaction, and learning theory within machine learning and natural language processing research. Further earlier topics include grounding, semantics, and reasoning, which sit among many topics centered on machine learning and natural language processing. The group also champions reproducibility through the creation of CodaLab Worksheets, a resource intended to support repeatable experimental practice in machine learning research.
Pranay Joshi Lab
The University of AlabamaPRanay Joshi · Assistant Professor
An illustrative lab studying how people interpret AI uncertainty and how reliable systems support better decisions.
Sameer Singh Research Group
University of California, IrvineSameer Singh · Professor of Computer Science
How can machine learning models be made robust and interpretable? Sameer Singh is working primarily on the robustness and interpretability of machine learning algorithms and models that reason with text and structure for natural language processing. He was a postdoctoral researcher at the University of Washington and received his Ph.D. from the University of Massachusetts, Amherst. His group has received funding from Allen Institute for AI, Amazon, NSF, DARPA, Adobe Research, Hasso Plattner Institute, NEC, Base 11, and FICO. Sameer has published extensively at machine learning and natural language processing venues.
Soroush Vosoughi Research Group
Dartmouth CollegeSoroush Vosoughi · Associate Professor of Computer Science
How can computational models of cognition be built when language is the primary conduit? Soroush Vosoughi leads the Minds, Machine, and Society group in the Department of Computer Science. The group's research primarily focuses on natural language processing and machine learning. It studies large language models to understand and mitigate their anti-social tendencies such as promoting stereotypes, spreading toxicity, and misalignment with human values. It also builds computational tools using LLMs and classical NLP to study phenomena such as political polarization, bias, propaganda, rumors, and hate speech.
Thorsten Joachims Research Group
Cornell UniversityThorsten Joachims · Jacob Gould Schurman Professor of Computer Science and Information Science
Clicks, choices, and text edits contain learning signals—but they also carry bias. Thorsten Joachims and his students develop machine-learning methods for using that feedback in search, recommendation, education, and language models. Their work connects counterfactual evaluation, policy learning, ranking, and personalization. POTEC tackles learning from logged decisions when the action space is large; coactive learning studies how a user’s edits can help personalize an LLM without requiring a perfect reference answer. This faculty-led research group offers a route into both the theory of interactive learning and its application to human-centered systems.
Tom Goldstein Research Group
University of Maryland, College ParkTom Goldstein · Volpi-Cupal Endowed Professor
How can AI systems be built responsibly? Tom Goldstein is a Professor of Computer Science and Director of the Maryland Center for Machine Learning, with appointments in Applied Math and Electrical and Computer Engineering. His research draws on ideas from signal processing and applied mathematics. He works at the boundary between theory and systems, using mathematical foundations and efficient hardware to improve the training, fine-tuning, benchmarking, and robustness of large models for language, computer vision, and science. His work on watermarking language models received an Outstanding Paper Award at ICML 2023.
William W. Cohen Research Group
Carnegie Mellon UniversityWilliam W. Cohen · Professor
How can machines answer questions using neuro-symbolic reasoning? William Cohen is a Professor at Carnegie Mellon University in the Machine Learning Department. He also holds a 20%-time appointment as a Principal Scientist at Google, where he worked full-time between May 2018 and March 2024. His research interests include question answering, machine learning for NLP tasks, and neuro-symbolic reasoning, and he has a long-standing interest in statistical relational learning. He holds seven patents related to learning, discovery, information retrieval, and data integration, and is the author of more than 300 publications.
Showing 1–24 of 26 labs