An international perspective on ubiquitous computing and university education.
Erol Gelenbe's research interests cover packet network design, computer performance analysis, artificial neural networks and simulation with enhanced reality. He is the University Chair Professor of Electrical Engineering and Computer Science at the University of Central Florida, where he serves as Director of the School of EECS and Associate Dean of Engineering. Previously, he was Chair of Electrical and Computer Engineering at Duke University. His many honors include Fellow of the IEEE , Fellow of the ACM, Chevalier de l'Ordre National du Merite (1992) and Officier de l'Ordre National du Merite (France, 2001).
UBIQUITY: Let's start by talking about your background.
GELENBE: I was born in Istanbul, Turkey. I got my undergraduate degree at the Middle East Technical University, which is a school of engineering and science in Ankara, Turkey. In my senior year, I got involved in a research project designing fast switching magnetic core memory systems. I was fortunate to get a Fulbright fellowship to pay for grad school. So I came to the US and went to Brooklyn Poly.
UBIQUITY: Why Brooklyn Poly (the Polytechnical University of New York)?
GELENBE: The main reason I went to Brooklyn Poly is that the dean of engineering at Middle East Technical University was a Brooklyn Poly alumnus, and he encouraged many of us to go there. The secondary reason was that I was interested in electronics, and at the time, there was a good group working in that area at Brooklyn Poly. I got my masters and PhD degrees at Brooklyn Poly in a period of three years. That's a very fast summary of the initial part of my existence.
UBIQUITY: What did you do next?
GELENBE: As a Fulbright fellow, I had to move out of the US after five years, so I had two more years to stay. After I was at Brooklyn Poly, I was a faculty member at the University of Michigan. I spent two years there and then Michigan gave me a leave of absence. At that point, I had an offer to teach at my alma mater in Turkey. Unfortunately, at that time, in 1971, there was a military takeover of the government in Turkey. I was just married and my wife said, perhaps it's not such a good idea to go back there because it looks like the university is being messed up.
UBIQUITY: Was she right? Was the university messed up?
GELENBE: Let's put it this way. The university was viewed as more of a liberal place. And the military was less liberal. In retrospect I can say that many of my friends who did go back to Turkey during that period did have problems developing an academic career. I had an offer to go to France to do research at IRIA, the National Research Institute in Informatics and Automation, which is now known as INRIA. So we went there and things moved ahead rather nicely. I was given a group to manage and develop in the area of computer performance evaluation. When the time came to go back to Michigan, our son had been born and my wife didn't want to move again. We stayed in France for about 20 years.
UBIQUITY: You've had a lot of experience in both American and European institutions. How do they compare?
GELENBE: They are very different. American universities are what I'd call "total organizations" in that they handle, of course, teaching and research, but also student life. I think that's the major difference between European and American universities. European universities just view the teaching and the research. They're not social structures. They typically don't have the dorms, the infrastructure facilities, and the student unions that we're so used to seeing on American campuses.
UBIQUITY: Ah. Not even football teams.
GELENBE: Certainly not the football team. They may have clubs, but they don't run athletics as an institutional activity. From a student perspective, that, by itself, is a major difference. There are other differences. One other major difference is that research in many of the continental universities is something that faculty do on an individual basis. Of course, there's the notion of grants and so on. But the university doesn't have the same institutional stake in research that American universities do. The basic reason, you'll be surprised, is that there are hardly any indirect costs (i.e. "overhead") added to European research grants. An American university's research is financially rewarding to the university because it generates indirect costs. It also offsets the cost of graduate programs and the cost of faculty salaries. In Europe, you don't have that. So the European universities are not as institutionally motivated to see research happening. Some of them even consider it to be a humbug because it interferes with teaching. It also takes up lab space and faculty time.
UBIQUITY: Do the governments give direct research support?
GELENBE: They do give a certain amount. Sometimes, our colleagues in American universities have this fantasy that if they go to Europe, that there will be wonderful funding that will allow them to do interesting work and travel to Honolulu for conferences and so on, where it's just not that way. The other thing is that in the US the faculty get additional, what we call summer salary, which is basically an add-on to our salary. While in European universities, they don't have that. Salaries are computed on a 12-month basis, with nine months of teaching and three months of research. That's the general view. So the faculty don't have a strong incentive to get grants. Both the student life aspect and the research and grants aspect distinguish the continental European model from the American model. UBIQUITY: You mentioned that you have a son. Where did he go to school?
GELENBE: He used five years very efficiently. He got an undergraduate degree in engineering at Duke University in two years while I was on the faculty there. Then, he spent another two years getting an engineering degree at a French school. He then went to Columbia for a master's degree in operations research. So he basically went to three schools in five years. I was hoping that he'd go on to get a PhD, but he was lured off into hedge funds and things of that nature. Quite early on, he co-founded a company in London called Flytxt, which was just written up in "Red Herring" under the title "Flytxt in the Digital Soup."
UBIQUITY: One thing that jumps out on a current resume, such as yours or your son's, is looking at how international American universities have become.
GELENBE: That's right. Many people view that as a weakness in American universities' graduate programs.
UBIQUITY: Why would it be considered a weakness?
GELENBE: Because they think, well, we have all of these foreign students. Why don't we get Americans? That's one way of looking at it. Another way of looking at it is that the kids we attract into graduate studies in the US are the elite of the undergraduate populations in their respective countries. The US strongly benefits from this influx of talent. These kids have all gone through the main hurdles. I think American universities are, in this way, very fortunate. Of course, they're fortunate because their aim is to become fortunate. In other words, American universities have infrastructure. They have faculty. They have talent. They have ideas. So these talented young people come from all over the world. If American universities were not so successful to begin with, these people would not want to come here.
UBIQUITY: Let me ask you to venture a value judgment of sorts. Apart from the whole issue of the social infrastructure and the football games and cheerleaders and all that kind of thing, does one intellectual experience tend to be "better" than the other?
GELENBE: American universities are extremely varied in their nature, in their quality and in the things they stress. There is more opportunity for diversity, for choosing the level that corresponds to what you want to do and what you're capable of doing. I think that's a major strength. No one claims that Duke, Harvard and the University of Central Florida are the same. You may have some courses at UCF that are as good or perhaps better than some courses at Harvard. But we're talking about different kinds of institutions. I think this is a strength of the American system as a whole. Furthermore, within American universities, we cater to different levels and strengths. For instance, at the undergraduate level at UCF, we have an honors college. The kids who are in the honors college have SAT scores that are comparable to their peers at say, Duke. But this is not the whole university. We are, within one university, able to cater to different talents and different levels. That is a big strength of the American system, where we don't try to apply the same model and the same standard to everyone. Therefore, the very talented kid will find his or her way through the system and develop his or her talents and even transfer to another university if he or she is not satisfied. This possibility of mobility and diversity is a major strength of the system in the United States. In that respect, I'd say that the US system offers a better set of opportunities, including intellectual opportunities.
UBIQUITY: Compare it to your experience in Paris.
GELENBE: If you go to school in Paris, and you work your way through the system, you'll be able to go into classes that are taught by some of the best people in the world. So yes, you do have this possibility in Paris but it's not as well organized. Many French universities don't have catalogs, so you can't even find out which course is taught when. The whole infrastructure for the student is standard in the US. In Europe, it's the exception. I think that's why a lot of well-off people send their children to the US to school. The system here works better for the student.
UBIQUITY: What drew you to the University of Central Florida?
GELENBE: From France, I moved to Duke and was awarded an Endowed Chair. I was at the beginning of my second term as department chair there when we moved to UCF. There are basically two things that enticed my wife and myself here. One thing was that I was given the opportunity to build the department, including recruiting a significant number of people. This was just not possible at Duke. We were very limited in resources. When I joined the Electrical and Computer Engineering department at Duke, we had 20 faculty. When I left, we still had 20 faculty. At UCF, just this past year, I have recruited seven faculty members. In three years, I have recruited 15 faculty members. What I was offered to do here was more exciting. The second determining factor was that UCF also offered my wife, Deniz Arman Gelenbe, a position. She is a musician and Juilliard graduate. She was offered to be an artist-in-residence and an associate professor here at UCF.
UBIQUITY: What does she play?
GELENBE: She's a pianist and chamber musician. In fact, later this month she will be playing at Carnegie Hall with her Arman Trio. She has an active performing career, and also teaches and organizes university-based concert series, and so on.
UBIQUITY: Talking about building your own departments, what is your vision of what kind of computer science department is needed?
GELENBE: When you think of needs, you think of the needs of the school you're at. One characteristic of UCF is that we have a very large student body. We have more than one thousand students majoring in computer science. UCF has a very strong undergraduate program. We are typically number one in the Southeast region in the programming team competitions of ACM. We typically run second or third place in the United States in the programming team competitions, behind Cal Tech, and sometimes, MIT. Anyway, when we think about computer science, we can't say we're going to concentrate just on operating systems or we're going to just concentrate on networks. When you recruit, you can't be shortsighted about the kinds of people that you need because even if everyone is not going to teach the same amount of courses , and we don't expect them to, you want to have competence in many areas. I've always tried to recruit people with a balance of theory and experimentation. For instance, today, if I were recruiting a graph theorist, I would be interested in a person who is a graph theorist but also in the course of his or her research, actually develops systems or does experiments using graph theory. The experimental component today is very important, more so than in the past.
UBIQUITY: Why is the experimental component more important today?
GELENBE: One of the reasons is that 20 or 25 years ago, experimentation in computer science was very difficult, very slow. We didn't have good tools. Everything had to be developed from scratch. Today, if you have good ideas, these good ideas can be turned into an experiment very quickly. Furthermore, experimental equipment in computer science is relatively cheap, say compared to physics, chemistry or biology. Basically, if you're doing advanced biology, you're talking about half a million dollars of equipment just to get into the game for a research scientist or a faculty member. Computer science is probably 20 percent of that cost to get into the game. So we can afford to have an experimental setting. My vision is to recruit people who have fundamental capabilities to understand theory, who are perhaps also engaged in theory, but at the same time who have an experimental activity.
UBIQUITY: Beyond experimental, is there any interest in interdisciplinary kinds of work?
GELENBE: Yes, there is. With interdisciplinary work, you have to get both disciplines to collaborate. Many disciplines still have a fairly primitive view of what computer science is. In many cases, you get people who think they know computer science because they use Word or Windows or because they use e-mail or know how to manipulate the Web. They think that your interaction with them is going to be roughly at that level. Interdisciplinary collaboration has to be set at the right level so that everyone is making progress. It's not about transferring things from one place to the other, but actually making progress on interdisciplinary collaboration. I myself have been involved in quite a bit of it.
UBIQUITY: Tell about some of your experience in collaboration.
GELENBE: Over the last seven or eight years, I worked in computational neuroscience with psychologists and neuroscientists. I have a practical view of what the difficulties of such collaborations may be. It takes say, more effort to do something good when you're doing something interdisciplinary than if you just stick to your own area. It is a significant additional effort. It's probably harder for junior faculty to get involved in that and much easier for senior faculty, who take fewer personal risks in doing so. For instance, we have a certain number of interdisciplinary projects going on. The ones that seem to be working to the advantage of the faculty are those where senior faculty are involved. Junior faculty may get burned out on it.
UBIQUITY: Is that because of tenure considerations?
GELENBE: It's not just the sheer tenure consideration. It's what are you accomplishing? What is the actual output of your work? Have you helped a biologist do an experiment or have you helped a biologist run some programs? Have you made some progress in biology or have you yourself progressed in computer science? That's the real challenge.
UBIQUITY: Has computer science over the years, and I won't specify how many years, acquired new prominence in the university?
GELENBE: I'd say no it has not. When the administrators have a few million dollars to invest in something, they would rather invest in physics or chemistry or biology, rather than computer science. We see that happening all the time. I think there are two reasons for that. One is purely the mental framework. Most of the people who control universities today are not computer scientists. Stanford is one of the rare schools where the president is a computer scientist. There's the Silicon Valley effect. That is certainly not the case at say, Duke or Princeton. University administrators are often either from the humanities or from the basic sciences.
UBIQUITY: And what is the second reason?
GELENBE: The second reason is that computer science, contrary to materials science or physics, funds itself quite well through teaching. After all, we have large numbers of students who major in our discipline. So the university has to mechanically give us resources that correspond to that. Now, this may sound great but the net effect is that we don't get the resources to take chances at the research level. Any university allocates resources, at least partially, based on the number of students. If there are many majors in computer science and there are no majors in materials science, the chances are that computer science will get more base funds than material science. So you see, computer science departments have big budgets anyway because they have all these students. Therefore, they do not need any additional funds. The negative effect is that we don't have the money to address intellectual challenges. We get funds for real teaching related work. It's not like we get funds and then we can scratch our heads and ask what we're going to do with them? We use them for the teaching purpose for which they were given. The result is that we don't have the internal funds to invest in advanced ideas or to take the chances that are needed in any field.
UBIQUITY: As you look into the future, do you see this changing anytime soon?
GELENBE: Unfortunately, I don't see this changing in my time. It will certainly change later when more computer science or information technology people are in positions of responsibility. But very few are. Very few deans of engineering are computer scientists.
UBIQUITY: When do you think it will change, 10 years from now, 20 years from now?
GELENBE: I think that things will start changing in the 15- to 20-year framework. I don't see that really changing within 10 years.
UBIQUITY: As you look back over, let's say, the last 10 years, can you think of any developments that genuinely surprised you?
GELENBE: Well, I think the Internet, as a whole, because when I was starting my research career in the '70s, networking activity was viewed as some kind of curiosity. Today, networks are pervasive. And I'm sure that when we were doing these things initially, talking about datagrams and packets and protocols and so on, we had absolutely no idea that they would become so common. I think we're not done yet with the way networks are going to be even more pervasive. Certainly, in retrospect, it was a surprise, although, it was a gradual surprise. We went from step to step but it's also quite astonishing that a few basic concepts developed say, 25 or 30 years ago, have proved so resilient and have become universal. When I started off in computing in the late '60s, who'd have thought that we would be able to have such cheap computing devices available everywhere?
UBIQUITY: What do you think is the most important thing that could happen, either in academia or more generally in society, as far as the interplay between the world and computing?
GELENBE: There is the purely conceptual aspect, which is, what is changing in the discipline or how are we understanding more complex things? Then, there's the aspect of how does computer science interact with society? I think that the societal interaction will come through a complete integration of computing in everyday life: ubiquitous computing. There will be sensory systems in buildings, sensory systems in cars, and complete media walls in homes and schools. You'll have virtual representations of things that you're interested in. Your art will come to your feet, in a certain sensory media. Of course, a lot of garbage will also come to your feet. But information technology will offer a great opportunity for access to culture. That is probably, to me, the most significant development that will come along. The possibility of having lots of great books at home immediately in your virtual library, the possibility of having great art, the possibility of having live performances of the New York Philharmonic at home or at school in Orlando or Ankara while it plays at Lincoln Center, and so on. We will have real digital libraries at our fingertips. How do we teach people to take advantage of the best? Because, after all, everything is just data and it is education and culture that allow us to discern the best from the worst. There's very bad painting and then there's very good painting. The difference between the two is not the fact that you can get to the painting. The challenge will be to learn how to understand and discern what is important, what is relevant, what is esthetically significant. The printing press revolutionized knowledge and education; it enabled the scientific revolution and opened the door to the age of enlightenment. Similarly, ubiquitous information technology will enable a major paradigm shift in our access to culture, the arts, science and technology.
UBIQUITY: Do you think our educational institutions are ready for this challenge?
GELENBE: I don't know. I certainly hope so. I especially hope that we will all be able to recognize what is common to all mankind, and is of universal value. Let us hope for the best. If we learn how to appreciate and manage the new possibilities, humanity will reap unimaginable rewards.
A Ubiquity symposium is an organized debate around a proposition or point of view. It is a means to explore a complex issue from multiple perspectives. An early example of a symposium on teaching computer science appeared in Communications of the ACM (December 1989).
To organize a symposium, please read our guidelines.
Ubiquity Symposium: Big Data
- Big Data, Digitization, and Social Change (Opening Statement) by Jeffrey Johnson, Peter Denning, David Sousa-Rodrigues, Kemal A. Delic
- Big Data and the Attention Economy by Bernardo A. Huberman
- Big Data for Social Science Research by Mark Birkin
- Technology and Business Challenges of Big Data in the Digital Economy by Dave Penkler
- High Performance Synthetic Information Environments: An integrating architecture in the age of pervasive data and computing By Christopher L. Barrett, Jeffery Johnson, and Madhav Marathe
- Developing an Open Source "Big Data" Cognitive Computing Platform by Michael Kowolenko and Mladen Vouk
- When Good Machine Learning Leads to Bad Cyber Security by Tegjyot Singh Sethi and Mehmed Kantardzic
- Corporate Security is a Big Data Problem by Louisa Saunier and Kemal Delic
- Big Data: Business, technology, education, and science by Jeffrey Johnson, Luca Tesei, Marco Piangerelli, Emanuela Merelli, Riccardo Paci, Nenad Stojanovic, Paulo Leitão, José Barbosa, and Marco Amador
- Big Data or Big Brother? That is the question now (Closing Statement) by Jeffrey Johnson, Peter Denning, David Sousa-Rodrigues, Kemal A. Delic