UMBC’s Prof. Cynthia Matuszek receives NSF award for robot language acquisition

Professor Cynthia Matuszek has received a research award from the National Science Foundation to improve human-robot interactions by enabling them to understand the world from natural language in order to take instructions and learn about their environment naturally and intuitively. The two-year award, Joint Models of Language and Context for Robotic Language Acquisition, will support Dr. Matuszsek’s Interactive Robotics and Language Lab, which focuses on how robots can flexibly learn from interactions with people and environments.

As robots become smaller, less expensive, and more capable, they are able to perform an increasing variety of tasks, leading to revolutionary improvements in domains such as automobile safety and manufacturing. However, their inflexibility makes them hard to deploy in human-centric environments, such as homes and schools, where their tasks and environments are constantly changing. Meanwhile, learning to understand language about the physical world is a growing research area in both robotics and natural language processing. The core problem her research addresses is how the meanings of words are grounded in the noisy, perceptual world in which a robot operates.

The ability for robots to follow spoken or written directions reduces the adoption barrier for robots in domains such as assistive technology, education, and caretaking, where interactions with non-specialists are crucial. Such robots have the potential to ultimately improve autonomy and independence for populations such as aging-in-place elders; for example, a manipulator arm that can learn from a user’s explanation how to handle food or open novel containers would directly affect the independence of persons with dexterity concerns such as advanced arthritis.

Matuszek’s research will investigate how linguistic and perceptual models can be expanded during interaction, allowing robots to understand novel language about unanticipated domains. In particular, the focus is on developing new learning approaches that correctly induce joint models of language and perception, building data-driven language models that add new semantic representations over time. The work will combines semantic parser learning, which provides a distribution over possible interpretations of language, with perceptual representations of the underlying world. New concepts will be added on the fly as new words and new perceptual data are encountered, and a semantically meaningful model can be trained by maximizing the expected likelihood of language and visual components. This integrated approach allows for effective model updates with no explicit labeling of words or percepts. This approach will be combined with experiments on improving learning efficiency by incorporating active learning, leveraging a robot’s ability to ask questions about objects in the world.

Meet the Staff: Alex Hart

Name: Alex Hart

Educational Background: Bachelor’s degree in Accounting from the University of Maryland, College Park

Hometown: Baltimore, MD (Go O’s and Ravens!)

Current role: As an Accountant I, Alex provides business services support to the CSEE department in the areas of contracts and grants/projects, which includes account monitoring, financial reporting, projections, reconciliations, etc. She also provides backup support for payroll, and she is the property custodian of inventory for CSEE.

Favorite thing about UMBC: “Without a doubt, my favorite thing about UMBC is the people here. I have met a lot of different people who have provided me with a wealth of knowledge since I started working here just a year ago. Everyone has been very inclusive and helpful!”

Students should ask me about: “Students can ask me anything, but maybe about the college experience, since I’m still a recent graduate.”

Alex is originally from Baltimore, MD. She joined CSEE’s Department in February of 2016. She attended UMBC for her first two years of college, then transferred to the University of Maryland College Park’s Robert H. Smith School of Business. She has a BS in Accounting from UMCP.

When not working, Alex loves cheering on the Terps in football and basketball. She also enjoys traveling to new places, cooking, practicing yoga, and reading.

UMBC computer scientists explain how AI can help translate legalese before online users click “agree”

 

Every day, people interact with large amounts of text online, including legal documents they might quickly skim and sign without full, careful review. In an article recently published in The Conversation, Karuna Joshi, research associate professor of computer science and electrical engineering, and Tim Finin, professor of computer science and electrical engineering, explain how artificial intelligence (AI) is helping to summarize lengthy and complex legalese so people can more easily understand terms of service and similar agreements before they click “accept” to access a new app or online service.

The legal documents that Joshi and Finin are working to summarize—terms of service, privacy policies, and user agreements—often accompany new online services, contests, apps, and subscriptions. “As computer science researchers, we are working on ways artificial intelligence algorithms could digest these massive texts and extract their meaning, presenting it in terms regular people can understand,” they explain.

Through their research, Joshi and Finin ask computers to break down the terms and conditions that regular users “agree” to or “accept.” To process the text, Joshi and Finin employ a range of AI technologies, including machine learning, knowledge representation, speech recognition, and human language comprehension.

Joshi and Finin have found that in many of the privacy policies people are prompted to review and accept online, there are sections that do not actually apply to the consumer or service provider. These sections of the agreements might, for example, “include rules for third parties…that people might not even know are involved in data storage or retrieval,” they note.

After examining these documents, the software Joshi and Finin have developed pinpoints specific items that people should be aware of when they are granting their consent or agreement—what they describe as “key information specifying the legal rights, obligations and prohibitions identified in the document.” In other words, the software takes in all that complex legal language, and then then presents just the most essential information in “clear, direct, human-readable statements,” making it much more feasible for users to understand what they are consenting to before they click “agree.”

Read “Teaching machines to understand — and summarize — text” in The Conversation to learn more about Joshi and Finin’s approach to making online legal documents more accessible through AI.

Adapted from a UMBC News article by Megan Hanks Banner image: Karuna Joshi. Photo by Marlayna Demond ’11 for UMBC.

Prof. Marie desJardins named by Forbes as one of 21 women who are advancing AI research

An article on Forbes’ site this week cites UMBC’s Professor Marie desJardins as one of 21 women who are advancing A.I. research. The article notes that artificial intelligence is “eating the world, transforming virtually every industry and function” and highlights women who are AI educators, researchers and business leaders who are driving the development and application of AI technology.

Professor desJardins joined the faculty at UMBC in 2001, after spending ten years as a research scientist in the Artificial Intelligence Center of SRI International in Menlo Park, California. She received her Ph.D. in computer science in 1991 from the University of California, Berkeley where her dissertation advanced autonomous learning systems in probabilistic domains.

The Forbes article states that

Marie desJardins has always been driven by broad, big-picture questions in AI rather than narrow technical applications. For her PhD dissertation at Berkeley, she worked on “goal-driven machine learning” where she designed methods an intelligent agent can use to figure out what and how to learn. As an Associate Dean and Professor at University of Maryland, Baltimore County (UMBC), desJardins has published over 120 scientific papers and won accolades for her teaching, but is equally proud of work she’s done with graduate students on self-organization and trust in multiagent systems.

When desJardins started her career, the AI and computing industry attracted more diverse, multi-disciplinary talent. Over time, she observed that conferences are “increasingly dominated with papers that focused almost exclusively on one subproblem (supervised classification learning) and much less welcoming of work in other subareas (active learning, goal-directed learning, applied learning, cognitive learning, etc),” which she is worried will exacerbate the diversity gap in AI.

“We are already seeing a reconsideration of more symbolic, representation-based approaches,” desJardins observes. “Ultimately I think that we will build more and more bridges between numerical approaches and symbolic approaches, and create layered architectures that take advantage of both.”

Her current research focuses on artificial intelligence, particularly machine learning, planning and decision making, and multi-agent systems. She has published over 125 scientific papers on these topics, and was recently named one of the “Ten AI Researchers to Follow on Twitter” by TechRepublic. At UMBC, she has been PI or co-PI on over $6,000,000 of external research funding, including a prestigious NSF CAREER Award, and has graduated 11 Ph.D. students and 25 M.S. students. She is particularly well known on campus and in her professional community for her commitment to student mentoring: she has been involved with the AAAI/SIGART Doctoral Consortium for the last 16 years and has worked with over 70 undergraduate researchers and four high school student interns. She was awarded the 2014 NCWIT Undergraduate Research Mentoring Award and the 2016 CRA Undergraduate Research Mentoring Award in recognition of her commitment to undergraduate research.

Prof. Ting Zhu receives NSF CAREER award to develop Internet of Things technology

 

The National Science Foundation (NSF) has awarded Ting Zhu, assistant professor of computer science and electrical engineering, its prestigious CAREER Award for his work to significantly improve the existing sensing capabilities of common technologies, such as cell phones. The five-year grant will total nearly $500,000.

“We congratulate Dr. Zhu on his NSF CAREER Award, an important recognition of his inventive work and what it may mean for other fields,” said Karl V. Steiner, vice president for research. “Dr. Zhu’s recognition adds to our growing list of outstanding young faculty recognized by peers and national funding agencies for their potential to advance science and technology.”

Ting Zhu, right, works with a student in the lab.

Zhu’s research may have impacts on the daily lives of people who use their cell phones to track activities, like exercise and sleep. He finds that although many mobile phones have activity monitoring systems built in, these capabilities are often not used to their fullest potential.

“The purpose of my research is to enable Internet-of-Things (IoT) devices to conduct accurate, efficient, and scalable N-way sensing,” explains Zhu. “This award will allow me to leverage sensing capabilities from different IoT devices to significantly improve people’s daily life for applications such as personal health monitoring, indoor localization, and smart home automation.”

To integrate research with education, Zhu will collaborate with his colleagues at UMBC to disseminate his research to local communities and recruit underrepresented groups. His work also has potential applications for enhancing the teaching of sensing technologies in classrooms, and his lab will provide opportunities for diverse UMBC students to complete hands-on research related to sensing technologies, as well as potentially exploring connections with virtual reality and 3D scanning.

In the last two decades, UMBC faculty have received 34 NSF CAREER awards. Additional UMBC faculty honored with CAREER awards so far in 2017 include Lee Blaney, assistant professor of chemical, biochemical and environmental engineering, for his work on water contamination, and Tinoosh Mohsenin, assistant professor of computer science and electrical engineering, for her work on energy efficient implementation of deep learning technologies and machine learning algorithms that are developed to function similarly to the brain.

Republished from UMBC News; Photos by Marlayna Demond ’11 for UMBC.

CS Ed Club’s Meet Your Prof. Series: Marie desJardins, Noon Mon May 8, ITE227

This semester, the CS Education Club has started a mini lecture series for students to interact with faculty outside of the classroom. They will have Dr. Marie desJardins as their next speaker. She will give an informal presentation followed by a discussion at 12:00 Noon on Monday May 8 in ITE 227. Light snacks and refreshments will be provided.

Dr. desJardins is the Associate Dean for Academic Affairs for the College of Engineering and Information Technology and a full Professor in the Computer Science and Electrical Engineering department. She has had an amazing career as a computer science researcher and educator. She has done research ranging from Artificial Intelligence to building a community of CS educators and improving CS education at the high school level in Maryland. She oversees the MAPLE lab, with a primary research interest in multi-agent intelligent systems. She has been an advocate for a better student experience as an administrator in COEIT, as well as the director of the Grand Challenges program at UMBC. Read more of her background on her website.

Please RSVP on myUMBC at the CS ED Club’s event page.

UMBC Prof. Tinoosh Mohsenin receives NSF CAREER Award for Deep Learning Technologies

Tinoosh Mohsenin, assistant professor of computer science and electrical engineering, has received a CAREER Award from the National Science Foundation (NSF) to advance her research on energy efficient implementation of deep learning technologies and machine learning algorithms that are developed to function similarly to the brain. Her award totals $475,104 over five years. Mohsenin’s research will enable those in medicine, intelligence, and environmental science to adapt the technology developed in her lab to their own work.

“We congratulate Dr. Mohsenin on her NSF CAREER Award, an important recognition of her groundbreaking work and the impact it will have on other fields,” said Karl V. Steiner, vice president for research. “This recognition of Dr. Mohsenin adds to our growing list of exceptional young faculty recognized by peers and funding agencies alike for the incredible potential their work has to move science and technology forward.”

The CAREER Award will support Mohsenin’s work creating solutions to both software and hardware issues with hardware implementation opportunities in her lab and across many industries. She is the director of the Energy Efficient High Performance Computing (EEHPC) Lab at UMBC. Mohsenin is particularly focused on energy efficiency, and emphasizes the importance of user-friendly, battery-powered and low-cost hardware implementation techniques for future computing.

Professor Tinoosh Mohsenin in her lab.

 

In the medical field, Mohsenin hopes her low power deep learning technology will help physicians and medical professionals detect seizures and cancer more quickly and accurately by improving the analysis of highly complex brain signal and image data, beyond what can be gleaned from today’s standard brain monitoring and analysis techniques. Mohsenin hopes her work will also help people with significant mobility limitations who use small multi-modal sensors on their tongue as well as other methods to maneuver wheelchairs or command other technologies. More complex algorithms and their efficient hardware implementation can notably improve the responsiveness of such technologies for users.

“I am very excited about this award as it allows me to take my research to the next level and help society find new computing techniques for smart wearable or mobile devices,” Mohsenin explains. “Current deep learning models have not been explored for power-constrained smart devices, and this research can potentially revolutionize several fields including healthcare, transportation, ecology, surveillance and public utilities.”

The award will provide her with opportunities to engage more UMBC students in STEM research, particularly among women and minority. She also looks forward to inspiring local middle and high school students to pursue engineering majors and careers.

Adam Page ‘12, computer engineering and Ph.D. ‘16, computer engineering, worked in Mohsenin’s lab on research that will be continued through the CAREER award.

In the last two decades, UMBC faculty have received 34 NSF CAREER awards. Additional UMBC faculty honored with CAREER awards so far in 2017 include Lee Blaney, assistant professor of chemical, biochemical and environmental engineering, for his work on water contamination, and Ting Zhu, assistant professor of computer science and electrical engineering, for his work to develop a networked system that can accommodate solutions for wireless communications, machine learning and data processing.

Adapted from a UMBC News article. Photos by Marlayna Demond ’11 for UMBC.

Wikileaks hack highlights importance of cyberdefense basics, UMBC experts write

The Central Intelligence Agency’s latest leak is the most recent major hack exposing information that could possibly compromise national security. In The Conversation, Anupam Joshi and Rick Forno, explain that this hack is a reminder of how cyberdefense strategies must be continually improved to ensure sensitive information is protected.

Joshi is a professor and chair of the department of computer science and electrical engineering and director of UMBC’s Center for Cybersecurity, and Forno is the assistant director of the UMBC Center for Cybersecurity and director of UMBC’s graduate program in cybersecurity. Their latest article has been republished by media across the globe and has been read more than 20,000 times.

“This round of leaks, of documents dating from 2013 to 2016,…reinforces perhaps the most troubling piece of information we already know: Individuals and the government itself must step up cyberdefense efforts to protect sensitive information,” write Joshi and Forno.

They ask readers to consider the risk to security and privacy compared with the benefits and convenience of modern technologies. “As citizens, we must decide what level of risk we — as a nation, a society and as individuals — are willing to face when using internet-connected products.”

Any electronic device connected to the internet is susceptible to a cyber attack, Joshi and Forno go on to explain, noting, “It’s not necessarily a good idea to have always-on and network-enabled microphones or cameras in every room of the house.”

Joshi also spoke with CBS Baltimore about how hacks can impact technologies consumers use every day, such as cars that now feature high tech navigation and entertainment systems. “The more electronic gizmos you have in your car, the newer the car you have, the more you’re connected to the network with your car, the greater the probability something can be done to your car,” he explained. Still, he noted, a hacker would need to have advanced technical knowledge and, likely, close proximity to the car to carry out such an attack.

To ensure that sensitive information is protected, Joshi and Forno say that focusing on “the mundane tasks of cyberdefense” is essential to maintaining security for everyone, from government to individuals, although they emphasize that no internet-connected technologies are immune to cyber hacks. Ultimately, they write, “Keeping others out of key systems is crucial to American national security, and to the proper function of our government, military and civilian systems.”

Read the full article in The Conversation, and watch the complete interview on CBS Baltimore.  Adapted from an article in UMBC News.

Prof. Gymama Slaughter on the body as a battery at Baltimore’s Light City festival

How can we begin to use our body as a power source? The same way we use a battery: by harnessing its chemical energy. As part of the annual Baltimore Light City Festival: A Festival of Light, Music and Innovations, Dr. Gymama Slaughter will present her research work on “The body as a battery – harnessing its chemical energy to power wearable and implantable sensors that diagnose and monitor diseases.” Dr. Slaughter will show how her team is converting the biochemical energy in blood sugar into electrical power, and how it is used to power wearable and implantable sensors.

The HealthLab@LightCity conference brings together innovators and leaders from Baltimore and across the nation to explore emerging technologies and innovative practices that have the potential to improve the quality of life and health outcomes for all people, here and around the world.

HealthLab@lightcity is presented by Kaiser Permanente and will be held 8:00am-6:00pm on Monday, 3 April 2017 at the IMET Columbus Center (701 E Pratt St, Baltimore, MD 21202) as part of Baltimore’s annual Light City Festival.

talk: Semantic Approach to Automating Big Data and Cloud, 12pm Mon 2/20

A Semantically Rich Approach to Automating Big Data and Cloud

Dr. Karuna Joshi
University of Maryland, Baltimore County

12:00pm Monday, 20 February 2017, ITE 325b, UMBC

With the explosion of Big Data and the growth of data science, there is an urgent need to automate the data lifecycle of generation, ingestion, analytics, knowledge extraction, and archival and deletion. With a promise of rapid provisioning, scalability and high computing capability, cloud based services are being adopted as the default computing environment for Big Data analytics.

To effectively manage their data on cloud, organizations need to continuously monitor the rules/constraints and performance metrics listed in a variety of legal contracts. However, these documents, like Service Level Agreements (SLA), privacy policy, regulatory documents, etc., are currently managed as plain text files meant principally for human consumption. Additionally, providers often define their own performance metrics for their services. These factors hinder the automation of steps of the data lifecycle, leading to inefficiencies in using the dynamic and elastic elements of the Data+Cloud ecosystem and require manual effort to monitor the service performance. Moreover, Cloud-based service providers are collecting large amounts of data about their consumers including Personally Identifiable Information (PII) like contact addresses, credit card details, bank account details, etc. They are offering customized service level agreements which indicate how such data will be handled. To see whether these agreements meet individual or corporate requirements, or comply with statutory constraints, currently involves significant human effort.

In this talk, we present the semantically rich approach that we have developed to automatically extract knowledge from large textual datasets, specially legal documents, using text analytics and Semantic Web technologies. We describe the OWL ontologies that we have developed, and the techniques to extract key terms and rules from textual legal documents. We will also illustrate application of our work in domains such as education, healthcare and cybersecurity.

Karuna P. Joshi is a Research Assistant Professor of Computer Science and Electrical Engineering at the University of Maryland, Baltimore County. Her research focuses on Data Science and Big Data Analytics, especially legal text analytics; knowledge representation and reasoning; privacy and security of Big Data and Cloud; and cloud enabled Health IT services. She has published over 30 papers, including in journals like IEEE Transactions on Service Computing and conferences like IEEE Big Data and IEEE CLOUD. Her research is supported by organizations like DoD, ONR, NIST, NSF, GE and IBM. She was also awarded the TEDCO MII award for exploring the commercialization of her research. She has been awarded the prestigious IBM PhD Fellowship. She also has over 15 years of industrial experience, primarily as an IT project manager. She worked at the International Monetary Fund for nearly a decade. Her managerial experience includes portfolio/program/project management across various domains. She received the MS and PhD degrees in Computer Science from UMBC and bachelor’s degree in Computer Engineering from the University of Mumbai, India.

1 2 3 12