Faculty Research Areas
Faculty Research Areas
Alan W Black I ProfessorSpeech Synthesis• Enabling computers to speak in any style or language
Speech-to-Speech Translation• Portable, real-time conversation cross-lingually
Spoken Dialog Systems• Making computers and machines communicate through speech
Website
Jeff Bigham I Associate ProfessorCrowd-Powered Conversational Assistants• Creating robust dialog systems using on-demand human computation• Crowd-powered assistants that automate themselves over time
Tools for Crowd Workers• Applying AI to amplify workers abilities, and help workers improve their skills and earnings
Dyslexia Detection and Intervention• Using linguistics and human-computer interaction measures to detect dyslexia• Building tools that help people with dyslexia read and write better
Website
Taylor Berg-Kirkpatrick I Assistant ProfessorUnsupervised Learning• Developing models and methods that can learn to understand aspects of human language
without relying on hand-labeled data
Structured Prediction• Building inference procedures that can efficiently and accurately produce complex outputs,
including both structured linguistic annotations (like parse trees) and natural language utterances(like summaries)
Analysis of Human Artifacts• Using automated methods to analyze and decipher structured human artifacts, including text
but also sources like piano music, images of early modern books (e.g., First Folio of Shakespeare),and cuneiform tablets
Website
Timo Baumann I Systems ScientistResponsive Speech Interaction• Incremental Processing for Spoken Dialog Systems• Conversational Dialog Beyond Task-driven Systems
Speech Signal Processing• Large-scale and Corpus-based Analysis• Speaker Likability and Speaker Traits• Prosody, Language and Multi-modal Integration
Website
Justine Cassell I Professor, LTI; Associate Dean for Technology Strategy & ImpactEmbodied Conversational Agents• Generating dialogue and nonverbal behavior for autonomous embodied agents
in augmented environments
Socially Aware Dialogue Systems• Deep learning approaches to recognizing, reasoning about, and generating social cues in dialogue
to increase user engagement and task success
Educational Technology• Dialogue-based systems for online and in-classroom learning, particularly for marginalized populations
Website
Jaime Carbonell I Director, LTIActive and Proactive Machine Learning
Transfer and Multitask Learning• Applied to deep neural nets
Low-Resource Language Analysis• Entity extraction and linking
Machine Learning in Computational Proteomics
Website
Jamie Callan I ProfessorSearch Engine Architecture• Distributed, federated, and vertical search• Use of knowledge assets and other semi-structured information• Structured and annotated documents
Text Mining and Analysis• Analysis of social media
The Lemur Project• Open-source search engine and text mining software• Large-scale web datasets• Search services that support the research community
Website
Ralf Brown I Principal Systems ScientistInformation Extraction• Text normalization
Language Identification• Identifying short texts in 1000+ languages
Digital Forensics• Text extraction• Reconstructing corrupted files
Website
Robert Frederking I Assoc. Dean, Doctoral Programs, SCS & Chair, Graduate Programs, LTIMachine Translation• Speech translation• Endangered language support
Natural Language Applications• Information extraction from text
Website
Scott Fahlman I Professor Emeritus, LTI & CSDSymbolic Knowledge Representation and Reasoning• Ongoing research on the open-source Scone knowledge-base system• Flexible, human-like planning, integrated with world-knowledge
(The focus is "good-enough planning", not optimal planning)
Knowledge-Based Natural Language Understanding and Generation• Going all the way from text or speech to a useful representation of the knowledge• Making use of context and background knowledge for disambiguation and filling in missing information
Incrementally Constructed Networks for Deep Learning• Looking again at some old ideas about gradually building up neural networks to fit the task at hand -
as a means to solve deep problems with much less guesswork and computation
Website
Maxine Eskenazi I Principal Systems ScientistMultidomain Dialogue Systems• Research to create intelligent agents able to respond to a variety of topics and assemble
diverse dialog systems
Distraction Detection• Deep learning to achieve early prediction of driver distraction using multimodal information
Website
William Cohen I ProfessorInformation Extraction and Machine Reading• Extracting factual knowledge from text• Answering queries using textual information• Neural approaches to question-answering
Statistical Relational Learning• Probabilistic logic programs and query languages• Differentiable query languages• Neuro-symbolic learning and reasoning
Website
Eduard Hovy I Research ProfessorComputational Semantics for Natural Language Processing• Machine learning and deep neural methods to do information extraction, event detection, entity
detection and linking into Wikipedia, reference resolution, semantic typing, etc., in multiple languages
Social Media Analysis• Latent analysis methods to discover interesting features of events and users of social media
Information Extraction, Text Mining, and Summarization
Website
Alex Hauptmann I Research ProfessorMultimedia Analysis and Retrieval• Large scale analysis of internet and surveillance video• Multimedia for healthcare• Multimedia for human rights
Website
Matthias Grabmair I Systems ScientistNatural language processing of legal text• Document Retrieval• Question answering• Information extraction• Automatic summarization
Computational Models of Argumentation• Knowledge representation and inference in formal argument models• Argument schemes for case-based reasoning• Modeling purposive legal argumentation with values• Learning to weight arguments from data and predict outcomes
Website
Anatole Gershman I Distinguished Service ProfessorInformation Extraction from Text• Conceptual dependency parsing• Domain adaptation through interactive learning
Conversational Systems• Task-specific and thematic chat bots
Knowledge Representation and Reasoning Under Uncertainty• Fusion of prior and new information
Website
Louis-Philippe Morency I Assistant ProfessorHuman Communication Dynamics• Analyze, recognize and predict subtle human communicative behaviors during social interactions
Multimodal Machine Learning• Probabilistic modeling of acoustic, visual and verbal modalities• Learning the temporal contingency between modalities
Health Behavior Informatics• Technologies to support clinical practice during diagnosis and treatment of mental health disorders
Website
Teruko Mitamura I Research Professor
& Director, MIIS ProgramText Analysis• Events: Definition, detection, coreference and representation• Annotation on event and entity mentions and coreference-linking
Question Answering• Question answering for Japanese National Center for University Admissions• Various tests for other domains
Computer-Assisted Language Learning• Intelligent reading system for English (Smart Reader)
Website
Florian Metze I Associate Research ProfessorSpeech and Audio Recognition and Understanding• End-to-end speech processing using Deep Learning• Multi-modal training, grounding, and adaptation (e.g., audio-visual ASR, children’s speech)• Low-resource speech recognition and spoken term detection• Language-independent speech recognition (e.g., using articulatory features)• Distant microphone recognition and information extraction from conversations• Audio event detection, sequence modeling of non-speech sounds• Recounting and summarization of video events (e.g., how-to-videos)• Extraction of para-linguistic information from speech (e.g., personality)
Website
Lori Levin I Research ProfessorMultilingual Natural Language Processing• Using linguistics and language typology to improve performance of corpus-based methods• Low resource languages
North American Computational Linguistics Olympiad (NACLO)• High school students learn about human language and computation by solving puzzles
Website
Carolyn P. Rosé I ProfessorText Mining/ Computational Sociolinguistics• Modeling social processes in discourse• Deep learning of rhetorical structure• Social Media Analysis
Dialogue Agents •Reinforcement Learning for Adaptable Dialogue Agents•Dialogue agents for Learning, Health, and Wellbeing
Computer-Supported Collaborative Learning•Architectures for supporting online collaboration•Social Recommendation Algorithms•Learning in Massive Open Online Courses
Website
Bhiksha Raj I ProfessorAudio Analysis• Audio content analysis, with applications to acoustic intelligence, surveillance, content-based retrieval• Never ending learner of sound: a self-updating audio-content index for the web• Signal enhancement and separation algorithms
Privacy Preserving Signal Processing• Algorithms to preserve user privacy in speech & audio applications• Secure cloud computing techniques
Speech Processing• Robust speech recognition and core speech recognition technologies
Deep Neural Networks• Novel applications of deep networks and algorithms for practical deployment of deep networks
Website
Eric Nyberg I Professor and Director, MCDS ProgramOpen Advancement of Question Answering• Software architectures and algorithms for real-world QA applications (e.g., Jeopardy! Challenge,
BioASQ, LiveQA)
Machine Reading• Mixed-initiative information extraction, logical form creation and inference for automatic
knowledge-base construction in any domain
Interactive Analytic Learning• Reducing cost of training high-quality analytics for new domains
Website
Graham Neubig I Assistant ProfessorCore Natural Language Processing• Machine translation• Text analysis, language modeling, parsing• Multilingual NLP and cross-lingual transfer
Machine Learning for Natural Language Processing• Neural networks, un- and semi-supervised learning• Efficient training methods for NLP systems
Natural Language Interfaces to Computers• Generation of code from natural language and vice-versa
Website
Ravi Starzl I Assistant Teaching ProfessorInformation-driven Systems• Computational analysis and modeling of complex domains such as biological systems,
financial systems, and Internet topologies
Software Engineering and Machine Learning• Mathematical modeling, big data systems, and biology
Website
Rita Singh I Senior Systems ScientistSpeech and Audio Processing• Signal processing• Automatic speech recognition• AI-based techniques for intelligence gathering and analysis, crime investigation and forensics
Website
Alex Rudnicky I Research Professor EmeritusConversational Systems• Proactive learning through spoken interaction between agents and humans• Agents that engage in social, non-goal directed interaction
Spoken Language Systems• Dialog system architectures and dialog management• Spoken language understanding and situational awareness• Speech recognition for interactive systems
Website
Roni Rosenfeld I ProfessorForecasting Epidemics• Developing the science and technology of epidemiological forecasting• Embedding this technology in tools for public and private decision making• Forecasting the timing and spread of Influenza and Dengue• Evaluation of forecasting accuracy
Information and Communication Technologies for Development (ICT4D)• Speech based, telephone based information services for low-literate people in developing countries• Viral entertainment as a vehicle for information dissemination
Machine Learning for Social Good (ML4SG)
Website
Alex Waibel I ProfessorSpeech-to-Speech Translation
Neural Network / Deep Learning and Language Processing
Machine Learning
Machine Translation
Speech Processing
Multimodal and Multimedia
Website
Yulia Tsvetkov I Assistant ProfessorMultilingual Natural Language Processing• Enabling language technologies for low-resource languages• Integrating rich linguistic knowledge into deep learning models
Machine Learning for Social Good• Identification of bias in online conversations; debiasing• Identification of manipulation and propaganda in media
Modeling Effective Communication• Identifying effective negotiation strategies• Enabling style transfer
Website
Anthony Tomasic I Senior Systems ScientistSelf-driving Databases• Self-driving databases to remove most of the duties of a database administrator
via machine learning analysis of workload
Language Production in Children• Development of new systems that promote language production in children
via speech recognition and user experience design
Crowdsourcing in Transportation• Application of crowdsourcing techniques to problems in the transportation domain
Website
Richard Stern I ProfessorRobust Speech Recognition• Physiologically- and perceptually-motivated feature extraction• Microphone-array processing• Application of deep learning approaches
Auditory Perception• Spatial hearing• Application of robust recognition techniques to the hearing impaired
Music Information Retrieval• Score following and musical training• Content analysis of musical performance
Website
Yiming Yang I ProfessorGraph-based Multi-Relational Learning• Unified framework and algorithms for joint inference over multi-source heterogeneous graphs of data
Time Series Forecasting and Anomaly Detection• Over massive heterogeneous sensor networks• Collaborating with Boeing and DOE
Cross-Domain and Cross-Language Transfer Learning• With topic identification and event summarization
Website
Eric Xing I ProfessorMachine Learning and Statistical Methodology• Solving problems involving automated learning, reasoning, and decision-making
Large-scale Computational System and Architecture• High-dimensional, multimodal, and dynamic possible worlds in artificial, biological, and social systems
Website