Dilek Hakkani-Tür: Faculty

Dilek Hakkani-Tür: Faculty

Dilek Hakkani-Tür is a distinguished visiting instructor at UC Santa Cruz and a senior principal scientist at Amazon Alexa AI. She brings deep Natural Language Processing expertise and practical experience to the program.

Dr. Hakkani-Tür received an MS (1996) and later her PhD in computer engineering from Bilkent University (2000). She also holds a BS from Middle East Technical University (1994). As a researcher, she has been granted over 70 patents, has co-authored more than 200 papers in natural language and speech processing, and won dozens of awards.

“My background is as a computer scientist,” Dr. Hakkani-Tür said. “I did my Ph.D. in computational linguistics, working with language modeling, and then became a colleague of [Program Director] Marilyn Walker’s at AT&T. Later I worked for Microsoft Cortana and am now on Amazon’s Alexa team.”

She is the recipient of three best paper awards for her work on active learning for dialogue systems, from IEEE Signal Processing Society, ISCA, and EURASIP. She served as an associate editor for IEEE Transactions on Audio, Speech and Language Processing (2005-2008), a member of the IEEE Speech and Language Technical Committee (2009-2014), an area editor for speech and language processing for Elsevier’s Digital Signal Processing Journal and IEEE Signal Processing Letters (2011-2013), and served on the ISCA Advisory Council (2015-2019). She is the Editor-in-Chief of the IEEE/ACM Transactions on Audio, Speech and Language Processing, and a fellow of the IEEE (2014) and ISCA (2014).

She foresees plenty of work for students with strong computer science and Natural Language Processing experience.

“It’s quite a challenging field. We’re still far behind humans in terms of the ability to understand language so over the next ten to fifteen years as more NLP applications are developed and the technology improves there will be lots of work for our graduates to do.”

There has been lots of interest in the program from industry. She foresees NLP agents being developed for reading and interpreting customer reviews, interacting with customers, parsing and producing documentation, and translating among many other potential uses. She’s excited to be joining UC Santa Cruz.

“The Santa Clara campus is in the core of Silicon Valley, which is the place to be if you want to work in the area,” she said. “UC Santa Cruz has always had Natural Language Processing and I was very impressed with the students I’d met, so it’s really exciting to get a chance to impact and influence these fresh minds entering the discipline.”

Ananth Sankar: Advisory Board

Ananth Sankar: Advisory Board

Dr. Sankar is a Principal Staff Engineer at LinkedIn, where he works on machine learning problems in the areas of Natural Language Processing (NLP) and multimedia processing. This work feeds into LinkedIn’s search and recommendation systems, as well as content classification.

During his career, Dr. Sankar has worked in the areas of speech recognition, natural language processing, display power management and search and recommendation systems, the common denominator being the development of innovative machine learning algorithms targeted for real applications.

Before joining LinkedIn in 2018, he held senior research and leadership positions at Google, Microsoft, Cisco, Yahoo!, Nuance Communications, SRI International, AT&T Bell Labs, and also founded a startup called Moxair, Inc. He has given several invited talks, has over 50 refereed publications, and holds 11 patents.

At UC Santa Cruz he is advising the NLP MS program. In the program, he has guest lectured and, as an advisor, brings the perspective of someone who frequently hires engineers that do applied research and development in the areas of NLP and machine learning. The program is also a chance for Dr. Sankar to experience his fondness for academia. He began his career as a lecturer at the University of Pune in India and has taught various NLP and deep-learning related classes at Stanford, UCLA, and USF, and guided a PhD student at Stanford.

Dr. Sankar holds a PhD from Rutgers University (1991), and an MS from SUNY Stony Brook, both in Electrical and Computer Engineering. He received his B.Tech in Electrical Engineering from I.I.T. Kanpur, India, in 1986.

When he isn’t working on cutting edge research, Dr. Sankar enjoys music. He’s an amateur guitar player and singer, and is fond of bands like The Beatles, James Taylor, and Pink Floyd.

Yi Zhang: Faculty

Yi Zhang: Faculty

Department: Computer Science and Engineering
Graduate Institution: Carnegie Mellon University

Dr. Zhang’s career straddles two worlds. In the first, she’s an academic, a professor of Computer Science and Engineering at the Baskin School of Engineering at UC Santa Cruz. In the second, she’s an entrepreneur, a co-founder of Rul.ai, an award-winning conversation computing platform that landed a spot on Forbes 50 Most Promising AI Companies list. She has also served as a consultant or technical adviser for several large companies and startups.

She received her Ph.D. from the Language Technologies Institute at the School of Computer Science at Carnegie Mellon University. She’s been at the forefront of Natural Language Processing (NLP) research for decades. Her research focuses on “two aspects, one is the impact of the problem, is the problem important to work out? Second is intellectual merit, is that problem worth research or a solution?” Her interests include large-scale information retrieval, personalized recommender systems, conversational AI, and applied machine learning. Among her many accolades: an NSF Faculty Early Career Award in 2010, an Air Force Research Young Investigator Award in 2008, and Best Paper Award at ACM SIGIR in 2002.

She teaches data mining and information retrieval courses in UC Santa Cruz’s Computer Science program and runs the Information Retrieval and Knowledge Management lab at UC Santa Cruz. Many of Zhang’s Ph.D. and Master’s students have gone on to work with NLP in industry: “The NLP program provides students with experience that people usually encounter in their first few years in industry. Being located in Silicon Valley and receiving instruction from actual industry experts are additional benefits for my students.” Her students tend to be multifaceted, but all enjoy becoming enmeshed in doing research and building solutions to solve challenging real world problems. Given the demand for NLP researchers and engineers, she’s delighted that many graduates are discovering a rewarding career waiting for them after leaving academia.

Like many of her colleagues, Zhang enjoys traveling with her family—particularly to places with interesting history and architecture. She also likes playing cards, skiing, and water activities.

Website: https://www.soe.ucsc.edu/people/yiz

Kaleen Shrestha

Kaleen Shrestha

Department: Computer Science and Engineering
Degree Program: Natural Language Processing MS
Hometown: San Jose, CA
Undergraduate Institution: University of California, Santa Cruz
Graduate Institution: University of California, Santa Cruz

Kaleen Shrestha is a member of the inaugural class of Baskin Engineering’s new M.S. in Natural Language Processing (NLP) program. She is also an undergraduate alumna of UC Santa Cruz, where she majored in computer science.

What made you decide to attend UCSC’s NLP Program?

I attended UCSC as an undergraduate and received my bachelor’s in computer science. I was fascinated by AI and wanted to explore the topics in that field some more, and so I decided to apply to graduate school. I was primarily interested in the ethical use of AI: understanding its implications and using it consciously.

When I applied to UCSC, I actually applied to the computer science and engineering MS program. Professor Marilyn Walker reached out to me through email because I had expressed interest in NLP in my application and talked to me about the NLP program. I got interested because the program was new and I thought it would be flexible and practical, and it showed a lot of room for growth.

Tell us more about your interest in NLP and how it has changed since you started the program.

I was curious about what interesting problems exist in NLP when joining the program. I had read a couple of interesting articles about NLP and had taken an introductory undergraduate course; in particular, an article that helped spark my interest in NLP was about gender bias in data and how it was amplified by the language model that was trained on that data. At that time I was taking Professor Lise Getoor’s new Ethics and Algorithms class and got interested in ethics in AI. I thought to myself, wow, things like this pop up in NLP and I want to work on this.

Now that I am a part of the program, I would say the courses really opened my eyes to the challenges in dealing with natural language. For example, in a linguistics class, we were learning about pragmatics and semantics which helped me realize that natural language is inherently biased on the individual and corpus level. Individually, there can be various styles for communicating a topic. On the corpus level, there are things like reporting bias, where the data available are only generated by certain types of people, for example. So this leads to problems in machine learning where models are dependent on patterns found in the data. So, I’m really glad to have found such an interesting subfield of AI where I can apply my interests in ethical data science.

What do you like about UCSC and the NLP Program?

Something that really stands out to me is my peers and fellow classmates. They have a great attitude towards learning. Everyone is very down to earth and humble when it comes to figuring out a challenging problem. This learning atmosphere is something that I didn’t see during my time in high school, and it was something that I really enjoyed during my time as an undergraduate at UCSC.

The question-asking and humble learning attitude was something that was really important for my learning and so that helped me choose UCSC for graduate school. In the NLP program, the professors are all very eager to help and open to answering questions. Since this is a new program, there are new situations that we have to adapt to and learn from, which I feel prepares us well for future endeavors in research and industry.

What would you like to do after you complete your NLP degree?

I want to ground myself in a practical application, that is something I really want to strive for and focus on. This program is really practical and I want to find a problem in industry to work on. I want to find a job and if I find an interesting problem, I would work on it in a PhD program or research.

What are some hobbies or fun facts about yourself?

I like gardening, and I recently propagated my avocado plant out of a seed. It took a whopping 6 months to get it to sprout, but now it is more than a foot tall! I also love hiking. At the UCSC campus, just walking to class is a hike.

Fun fact: my first name is made up completely, the spelling, everything! However, I found out that my name is the name of a well-established suburb in Australia. My name apparently also means water in an indigenous Australian language.

LinkedIn:
https://www.linkedin.com/in/kaleen-shrestha

Adam Fidler

Adam Fidler

Degree Program: Natural Language Processing M.S.
Undergraduate Institution: Brigham Young University, Utah Valley University
Department: Computer Science and Engineering
Hometown: Salt Lake City, Utah

Adam Fidler is a member of the inaugural class of Baskin Engineering’s new M.S. in Natural Language Processing (NLP). He spoke with us about what he likes most about the program so far and his aspirations for a career in NLP.

How did you discover your interest in Natural Language Processing?

I grew up with computers, so choosing to go in that direction was pretty easy. But after spending two years in Japan I realized I love languages too and wanted to see if there was an overlap between linguistics and computer science. After doing some research I found natural language processing (NLP) and realized it was something I wanted to do.

Why did you choose UCSC?

One of my NLP professors heard about the program from a linguistics mailing list. I applied on a whim, and I was happy to get in!

My undergrad prepared me pretty well for the things we’re talking about. I managed to essentially combine linguistics, computer science and applied mathematics into one major and it seems like it was good preparation for what I’m studying now.

What do you think of the program?

So far I like it. I think there will be plenty of networking opportunities, which is great because so many of our faculty members are in the industry right now.

My favorite class so far is machine learning. As an applied math student, I did a lot of machine learning. So the class means taking those things I learned about and applying them to linguistics.

I’m really looking forward to doing projects for our classes.

What are your plans after you graduate?

I love digital assistants like Siri and Alexa. Being able to work on something like that would be really neat. Everyone is using them… so I’d love to make them even better. I see them becoming much better at more generalized tasks. This is kind of a silly example, but I think of J.A.R.V.I.S. from The Avengers (Iron Man’s artificial intelligence) as the ideal. It would be wonderful to have something that generalized. We’re still a ways off from something so conversational and human-like. A lot of progress has been made and I’m happy to be pushing the limits of what we can do.

As for a career, academia has always intrigued me. I was a research assistant at BYU. It was a ton of fun and I love teaching, but I also really like the idea of going into industry and working for a startup, big tech company, or really anywhere that I could use the skills I pick up here because NLP is something I love.

What are some hobbies or fun facts about yourself?

I like hanging out with people, so of course these lockdowns have been driving me nuts. I hesitate to call math a hobby, but I think of it as one of my hobbies.

I took a course in Chaos Theory from the professor I worked for. You could say it’s the idea of studying how small perturbations affect something in the end, like the weather. It’s a chaotic, dynamical system we can model, but if you change the input ever so slightly, it can cause totally different results.

I spend more hours than is probably productive watching Youtube videos of math and math lectures.