Charles L. Isbell, Jr.
Executive Associate Dean
School of Interactive
Constellations Center for Equity in Computing
College of Computing
Georgia Institute of Technology
What do I do?
Although I tend to focus on statistical machine learning, my research passion is actually artificial intelligence. I like to build large integrated systems, so I also tend to spend a great deal of my time doing research on autonomous agents, interactive entertainment, some aspects of HCI, software engineering and even programming languages.
I think of my field as interactive artificial intelligence. My fundamental research goal is to understand how to build autonomous agents that must live and interact with large numbers of other intelligent agents, some of whom may be human. Progress towards this goal means that we can build artificial systems that work with humans to accomplish tasks more effectively; can be more robust to changes in environment, relationships, and goals; and can better co-exist with humans as long-lived partners.
After thinking about this problem for a number of years, I've decided that the central technical issues here are: adaptive modeling, especially activity discovery (as distinct from activity recognition); and scalable interaction, including coordination and influence. Further, I have come to believe that as a practical matter, it is necessary to build development environments that support rapid development, and so I have begun thinking seriously about authorial tools, including adaptive programming languages, domain-specific example-driven development, socially-guided machine learning, and corresponding issues in software engineering.
All in all, I believe that there are many opportunities in this space, and it should interest anyone who cares about any of the areas I mention above. I am building a research group to bear this out. To that end, I have found The Laboratory for Interactive Artificial Intelligence and the pfunk research group. Our research goal is to develop methodologies for building persistent, adaptive, collaborative, and believable agents that must live with other similar agents, including humans.
It is also worth pointing out that I have developed a strong interest in (re)defining Computing as a separate and vivid discipline. This has most obviously manifested itself in my efforts at curricular development and reform. Possibly related to these efforts, I was the Associate Dean of Academic Affairs in the College for a number of years before becoming the Senior Associate Dean and now the Executive Associate Dean (though I still retain my role in Academic Affairs while overseeing a lot of the operations of the College). Definitely related to these efforts, I have put a great deal of energy into thinking about equity in computing at all levels, with an eye toward the professoriate. I don't have dozens of papers in this space, but I care about it as deeply as I do my AI research and all the rest that I do. Generally speaking, I split my time among my professor and dean selves because I think efforts around operationalizing the breadth of computing deserve as much intellectual energy and thought as any of our other academic efforts.
So, if you're a Tech student (grad or undergrad), or perhaps want to be a Tech student, and you're dazzled enough by my html and css authoring abilities to think that we might have a fruitful conversation, feel free to contact me. Let's talk. If you do decide to email me, please use your Ga Tech address; otherwise, you run the risk of falling into the spam folder. If you're otherwise trying to contact me, putting "1988" in the Subject line allows you to avoid my spam filters as well. I may take a while to get back to you, but I will eventually.
Where do I come from?
I claim that I'm from Atlanta. I even went to college in Atlanta, at what is now Ga Tech's College of Computing. After graduating in 1990, I found myself at MIT's AI Lab (now CSAIL) pursuing a PhD and learning all about machine learning. For my dissertation, I developed a novel algorithm for inferring sparse, multi-level structure from a large collection of electronically available text. This was done in an unsupervised way, using principles from statistics and information theory. My advisors were Rod Brooks and Paul Viola. They are the first in a long line of illustrious academic ancestors, including Galileo, Copernicus, Jacobi, and Hess (no Bayes, alas).
After graduation I ended up at AT&T Shannon Labs, where I spent many fruitful years working with a slew of luminaries in machine learning (who happen to be spectacularly wonderful folks to boot). I left the Labs in the spring of 2002 in what has become known as the Great AT&T Labs Diaspora. I spent that summer hanging out at the University of Pennsylvania working with some good friends and colleagues of mine (most notably Michael Kearns and Lawrence Saul) on a variety of fun stuff. Now I'm back at Ga Tech, doing the faculty thing and reveling in the ongoing irony that is faculty life. I was awarded tenure (you should see what my students did to my office) and not too long after promoted to Professor, so I've decided that the academy is, in fact, a perfectly fair and meritocratic institution.
Do I have a life?
Why, yes... yes I do... or did at any rate.
Although research is obviously its own reward, I have pursued a truly ridiculous slew of hobbies over the years. For example, I sometimes wrote Hip Hop Reviews; used to run The Annual New Jack Hip Hop Awards; ran the funky-music and hiphop mailing lists; and after more than 15 years of maintaining the This Week In Black History page and calendar (I believe it was one of the first online databases of its kind), passed it on to some interesting folks.
I sometimes interact with the physical world, doing things like racquetball, weight-lifting and ultimate frisbee. I have a family, too, or so they tell me.
There's a lot to some of this stuff. You might find it interesting if you look into it.