Prof. Nesime Tatbul: Database Management Systems

Diese Seite existiert nur auf Englisch!

Nesime Tatbul has been an Assistant Professor of Computer Science at ETH Zurich since January 2007. She received her B.S. and M.S. degrees in Computer Engineering from the Middle East Technical University (METU) in Ankara, Turkey. She then joined the Computer Science Department at Brown University, where she obtained her Ph.D. degree in 2006. During her graduate school years at Brown, she also worked as a research intern at the IBM Almaden Research Center, and as a consultant for the U.S. Army Research Institute of Environmental Medicine. Her research interests are in database systems, with a current focus on data stream processing and networked data management. An interview with Katja Abrahams.

August 2007

What is your main research area?

My main research area is database management systems. More specifically, I focus on performance and scalability aspects of newly-emerging data management applications. Lately, I have been working on large-scale data stream processing, as in the example of continuously generated financial stock tickers, or data generated by wireless sensor networks for real-time monitoring applications such as environmental monitoring, patient monitoring, or highway traffic monitoring.

What are currently the most challenging issues in database systems research?

The core focus of database systems has always been on high-performance processing and analysis of very large amounts of data. With this goal, come many issues on how data should be modeled and stored; what interfaces, languages, and algorithms should be designed to access that data; how all the different pieces should come together in a scalable, general-purpose system architecture; to name a few. New applications of course pose new challenges, and we have to continuously rethink our previous assumptions and design decisions.

Currently, managing large volumes of data streaming from various sources in real-time is one of the challenging topics. The processing must be fast since we would like to react to the data as soon as the corresponding real-world events happen. These events might be very complex to detect, may involve remembering the data history, or making predictions about the future data arrivals. The data itself may contain imperfections (inconsistencies, missing values, disorder, etc), and may require some kind of cleaning or model-fitting before we can make sense out of it.

Managing different types of data on the web is also one of the big challenges. As opposed to the classical database systems, web data is usually unstructured; requires effective search techniques when not so much is known in advance about both the data structures and the data sources; web users are generating lots of content (blogs, Wikipedia entries, YouTube videos, etc.) while they are also interacting and exchanging data in social networks. There is also a growing need to be able to integrate data from different sources in a flexible way, whether the data is your personal data on your laptop, or it is somewhere out there on the web; whether it is streaming, or generated on demand.

Preserving the privacy and security of the data, especially when so much about our personal lives is exposed to the others on the web, is also an important challenge.

Which courses are you teaching this semester and what will you teach next semester?

In this upcoming semester, I am co-teaching a seminar and a lab together with Donald Kossmann and Moira Norrie. The seminar is called "Advanced Topics in Information Systems". It is offered every Fall semester by the Information Systems Institute, covering a different research topic every year. This year’s theme is "Event Stream Processing". We will focus on continuous data streams that represent real-world events, how they can be modeled and processed. This is one of the hot research issues in information systems lately, and also interesting in that the industry research is going almost head to head with university research (which doesn’t happen very often). The "Information Systems Lab" is also offered every Fall semester, where students build real systems to solve a particular data management problem.

In the Spring semester, I am planning to teach a new lecture on "Networked Information Systems". In this lecture, I would like to cover various architectural models for distributed information systems, and investigate some classical data management problems in the modern networked settings of today such as wireless sensor networks and the web.

Are there any topics you would like to teach but just can't fit into the course catalog?

I think that our course catalog is quite flexible and extensible. It is also great that we have the opportunity to co-teach courses with colleagues in areas of common interest. In the near-future, I would like to teach courses also at the basic undergraduate level.

What do you like about teaching?

I like the interaction with the students. Students bring a fresh point of view, and they come up with original and creative questions and comments. Essentially, learning is a life-long process, and you can learn new things every time you teach the same course. For that, I encourage students to actively question the things that are being taught to them, don’t take anything for granted. I also like the two-way synergy between research and teaching.

Why did you become a computer scientist?

Computer Science is a relatively young and exciting field of science, there is a lot to explore. CS touches and has the power to positively influence almost every field of life, from education and health to environment, art and business, and so forth. You can make a big difference in the way people live their lives.

Do you think more women should be encouraged to study CS?

Yes, definitely! There is a lot that they can contribute to and they shouldn’t consider CS as a "man’s field". Women bring a different point of view to every field; CS is just one of them. Any field without women contribution would be severely incomplete. I think that role models are very important to encourage more women to join us. It certainly was a factor in my career so far, and I would be very happy if I could act as one myself.

What do you think of ETH as a research center?

It is without any doubt one of the top scientific institutions in the world. The resources and opportunities to make significant scientific contribution and impact are almost endless. Not only financially of course, but with the smart and highly motivated student body, and exceptional colleagues to collaborate with. People are the most important ingredient.

Is there anything special you would like to say to our students or future students?

I haven’t been here that long. I am really looking forward to meeting more students. I would like to encourage all students to look at the courses and the research projects that we offer, and come and talk to us if they are interested in working with us. My group, Advanced Data Management Systems (ADMS), has a brand-new web site: http://www.adms.ethz.ch, where more information is provided. Our doors are always wide-open for motivated people!