Research Labs

Lab Director: Wenjin Zhou
Location: Dandeneau Hall 412
Email: wenjin_zhou@uml.edu
Wenjin Zhou's personal website

The Biomedical Computing and Visualization Lab at UMass Lowell focus on analyzing data in Biomedicine for enabling data-driven decision making in medicine, health care, and brain science. Our group tackle data science with three research components:

1. Computational Solutions and Machine Learning - To understand how data is structured, analyzed and processed through data modeling, information retrieval and data-intensive computing.
2. Scientific Visualization - To explore how we can make sense of our big data sets combining multi-dimensional information obtained in various format and imaging modalities.
3. Interactive User interfaces - To discover how interaction with the data at multiple scales can help us connect the dots and can lead to new knowledge.

These three research components interact with one another to expand our knowledge in computer science, brain and health sciences. We believe that application-driven contributions to expand computer science to interdisciplinary research will become more and more important in exploring and understanding scientific and health-related problems.

Our current projects span the following research areas:

• Computational Drug Design and Repurpose: when quantum mechanics meets machine learning
• Brain Disease Biomarker: quantitative microstructure extraction from diffusion MRI
• Interactive Multidimensional Brain Visualization: understanding and connecting brain data at multiple scales
• For more information on this lab, or information on how to apply to work here, please visit the lab’s website or email the lab director.

• Curriculum Learning for Natural Language Processing: Deep neural networks can effectively tackle many NLP tasks, but they could be computationally expensive to train. How can we uncover the salient characteristics of these learners (networks) and their learning materials (training data) for effective representation of textual data and efficient training?
• Clinical Decision Support: Most medical information, from physician notes to referral letters to scientific articles, are locked in unstructured text and are not readily accessible. NLP and Machine Learning techniques offer potential means to support clinicians with evidence and insight extracted from such data. What are the best techniques to represent multimodal patient data and reference materials about diseases, triage patient applications, pinpoint disease-causing gene variants, and enhance clinical decision support systems with evidence?
• Social Media Surveillance: User generated content in social media present naturally occurring data that can be used to obtain low-cost and high-resolution views into population behavior. How can we develop online surveillance systems that can monitor population behavior at scale to detect (health-related) trends and (disease) outbreaks, and identify opportunities for decision making, resource allocation or intervention?

Lab Director: Farhad Pourkamali Anaraki
Location: Dandeneau Hall 408
Phone Number: 978-934-6126
Farhad Pourkamali Anaraki's website

Prof. Pourkamali and his research group focus on the foundational and statistical aspects of machine learning as well as a diverse range of practical applications. Notably, his research team works on developing scalable and principled techniques to analyze complex high-dimensional data in unsupervised scenarios (clustering), and when labeled data is limited (active learning). Moreover, Prof. Pourkamali conducts interdisciplinary research in various areas, such as structural engineering and materials science.

The CSDS Lab hires undergraduate, master’s and Ph.D. students! If a student wants to get involved with this lab, they should take Machine Learning. It is also important to have a basic knowledge of calculus, linear algebra, and programming in Python.

Research Papers

• Improved Fixed Rank Nystrom Approximation via QR-Decomposition
• Large Scale Sparse Subspace Clustering Using Landmarks
• Preconditioned Data Sparsification for Big Data with Applications to PCA and K-Means
• Matrix Completion for Cost-Reduction in Finite Element Simulations Under Hybrid Uncertainties

Lab Director: Fred Martin
Location: Dandeneau Hall 408
Engaging Computing Group website

The Engaging Computing Group develops and studies novel technologies for K-12 computer science education, with a focus on data science, virtual reality, and embedded devices.

The ECG Lab loves to hire undergraduate students, as well as MS and Ph.D. students. If you want to get involved with the ECG Lab, contact Director Fred Martin. You must have a handle on the basics of CS (Computing I-II) but aside from that, students have a lot of creative freedom on what they want to work on. There are a bunch of cool existing projects to work on, but Martin is always receptive to new ideas.

One of the lab’s longest running works is the iSENSE Project, a web system for collecting, visualizing, and sharing data, geared towards elementary and middle school students, which is funded by the National Science Foundation. The site was designed by the lab, and is currently maintained by lab members who also add new features.

Lab Director: Michelle Ichinco
Location: Dandeneau Hall 412
Email: michelle_ichinco@uml.edu
Human-Computer Interaction Group website

The Human-Computer Interaction Group at UMass Lowell focuses on the design and evaluation of tools to support learning, especially of programming. Through the study of people and their interactions with software and tools, we work to understand ways to help people learn more efficiently. We design, develop, and user test tools that aim to move the future of learning forward. The HCI group publishes in venues like CHI, VL/HCC, IDC, and ICER.

For more information on this lab, or information on how to apply to work here, please visit the Human-Computer Interaction Group website or email the lab director.

Lab Director: Holly Yanco
Location: Dandeneau Hall 409
Phone Number: 978-934-3385
Human-Robot Interaction Lab website

Research at the human-robot interaction lab focuses on human-robot interaction (HRI), which includes multi-touch computing, interface design, robot autonomy, trust, and evaluation methods. Application domains include assistive technology, telepresence, and urban search and rescue (USAR). The robotics lab also has an active K-12 community partnerships through programs such as Botball.

The HRI Lab hires a new wave of undergraduate immersive scholars every year, along with a few undergraduates who apply independently of the immersive scholars program. In addition to this, Holly accepts resumes each year for Masters and Ph.D. students who have a passion for robotics and designing successful interfaces and user studies. If you want to get involved with an undergraduate, you want to be comfortable with version control, working with Linux, and have some experience with object oriented programming. The lab does a lot of work with Robot Operating System (ROS), so exposure to this is a bonus. Some classes that will set you up for success in this class include the Fundamentals of Robotics course, Mobile Robotics I-II, and Yanco’s very own Human-Robot Interaction class.

For more information on Prof. Yanco, the robots and people who work in the lab, and the current research going on, visit Prof. Yanco’s personal website or the UMass Lowell Human-Robot Interaction Lab website.

Lab Director: Reza Ahmadzadeh
Location: Dandeneau Hall 413
Phone Number: 978-934-6082
Email: reza@cs.uml.edu
Persistent Autonomy and Robot Learning website

The Persistent Autonomy and Robot Learning (PeARL) research lab focuses on the development of autonomous robotic systems that operate effectively in complex and unstructured environments, adapt to human feedback, and learn from humans. Directed by Prof. Reza Ahmadzadeh, the lab’s research spans persistent autonomy, physical human-robot interaction, Reinforcement Learning, and Learning from Demonstrations.

For more information on who the PeARL lab hires, job openings, and past research projects, please visit the very informative lab website listed above.

Lab Director: Anna Rumshisky
Location: Dandeneau Hall 415
Email: arum@cs.uml.edu
Text Machine Lab website

The Text Machine Lab at UMass Lowell conducts research in machine learning applications for natural language processing, with a focus on deep learning methods. Our current projects span the following research areas:

• Computational lexical semantics, distributional analysis and machine learning techniques for representing lexical and sentence-level meaning
• Text-based temporal reasoning, argument mining
• Clinical natural language processing, including information extraction, and predictive modeling using clinical patient records
• Digital knowledge and sentiment tracking for digital humanities and social science

The Text Machine Lab generally hires undergraduate students and Ph.D. students; MS students are only hired in exceptional conditions (e.g. they have done NLP/ML before coming to UMass Lowell). If a student wants to get involved with this lab, they should take machine learning, and learn some PyTorch; Andrew Ng's online course or the machine learning course here at UMass Lowell are both good starting points. Once you reach out and express your interest to the lab director, you’ll typically receive some good exercises to code up in order to get up to speed.

For an extensive list of projects, papers, and students who have been products of this lab, please check out the lab’s website! The TwitterHawk project was led by an undergrad.

“One thing I can say is that the undergraduate students that join my lab get involved in research early. I believe there are plenty of very talented undergraduates at UMass Lowell who should be given a chance to participate in research and given a chance to contribute meaningfully -- and unfortunately, that often does not happen until they get to their senior year. I try to do the opposite and give people challenging things to work on early on. As a result, I have undergraduate RAs who by the time they graduate have had more research experience that many first or second year Ph.D. students. Typically, undergraduates in my lab publish papers at major conferences and come out well prepared -- the very first of my undergraduate research assistants ended up receiving the NSF Graduate Fellowship and he is currently finishing his Ph.D. at MIT EECS.”
     - Anna Rumshisky, Lab Director

Lab Director: Jie Wang
Location: Dandeneau Hall 408
Jie Wang's personal website

The group performs research in text mining algorithms and systems, document engineering, and data modeling and its applications. Some of the projects that we have worked on include intelligent text automation, document summary, domain extraction and statistical measurement in knowledge networks, and news classification.

The Text Mining and Document Engineering Group hires Ph.D. students and occasionally master’s students. However, they do plan on hiring undergraduates in the future.

Linear reading from the beginning to the end is the normal way of reading for the purpose of learning new knowledge. Is there a better way to read for understanding? Let’s think outside the box and turn linear reading to hierarchical reading. Imagine that we have access to an oracle that ranks the sentences of the given text according to their importance, allowing us to read blocks of sentences one at a time in descending order of importance, focus on the most important block of sentences, and read subsequent blocks to strengthen understandings of earlier blocks. Moreover, the oracle also generates questions and evaluates answers. Dooyeed is an AI-assisted text mining system to facilitate reading for understanding to meet the AEE requirements: Accurate and Efficient for the oracle and Effective for the reader. Dooyeed outperforms each individual human judge over the SummBank benchmarks and compares favorably with the combined rankings of all judges. Dooyeed currently supports two major languages: English and Chinese. This invention is made by Jie Wang and assisted by his Ph.D. students, Students who have participated in the research and development of Dooyeed include Cheng Zhang, Hao Zhang, Changfeng Yu, Yicheng Sun, and You Zhou. Eola Solutions, Inc. provided financial support.

The Text Mining group has graduated a number of Ph.D. students and MS students, including Yiqi Bai (now at Facebook Seattle), Ming Jia (Facebook Seattle), Liqun Shao (Microsoft Cambridge), Jingwen Wang (Elizabethtown College), and Wenjing Yang (Dell EMC).