COSMOS, in this edition of Cosmographer Corner, highlights the work of former University of Arkansas at Little Rock graduate and Cosmographer Dr. Zachary Stine. We are extremely proud of Dr. Stine’s accomplishments!
Dr. Stine—who is now an assistant professor of computer science at the University of Central Arkansas (UCA)—started his graduate education at UA Little Rock in 2012, studying computer science as a post-baccalaureate. Dr. Stine started his PhD under Dr. Agarwal’s supervision and joined COSMOS as a graduate research assistant in 2016. He received his PhD in computer & information sciences in 2021, and worked as a postdoc at COSMOS. We interviewed Dr. Stine on where his career is now and what his work at COSMOS entailed, with his responses below.
How did COSMOS fit into your university/secondary education career? How did you come across COSMOS, and what were you studying when you joined COSMOS?
I was what’s called a post-bac student: I had finished my bachelor’s degree in an unrelated field, which was religious studies (the secular study of human religion). I knew I wanted to do a master’s program, particularly computer science. I ended up taking around 42 hours of coursework while I was working for a health insurance company. They were paying for me to take undergraduate classes. And so I took all of the core computer science curriculum, and decided OK, I want to do the master’s program. That’s when I fell in with Dr. Agarwal, when I was making that transition to the graduate program. When first studying computer science, I was interested in network science and social networks, and had been introduced to Dr. Agarwal through those classes. I was a student just trying to figure out which master’s program I wanted to do. After realizing that Dr. Agarwal was looking at the sorts of questions that I was interested in, I jumped at the chance to work with him, eventually deciding to just go ahead and do a PhD.
At first I wasn’t doing full-time classes, and wasn’t a full-time research assistant at that point. But eventually I decided to go ahead and take the plunge, leaving behind my job to become a research assistant with Dr. Agarwal.
But there was no COSMOS when I joined—it was simply Dr. Agarwal’s lab, at that point. I think it was a year into my work with Dr. Agarwal that COSMOS came to be—I remember we had a vote on different candidate names, before it was officially COSMOS. So, this was “proto” COSMOS.
How would you describe the “research pipeline” that you worked on while at COSMOS? In other words, what was the specific area in which you researched?
I worked on a variety of projects that primarily involved text data. When I first started, I was doing agent-based models of opinion dynamics and then moved into text-related things. One project I was heavily involved in was on Ukraine and the Ukrainian government. We also had a variety of one-off projects that I was doing as part of my dissertation work, which involved comparing how people talk about different things. That used a lot of Reddit data, which was unique because at the time everybody in the social media field was focused on Twitter, almost to the exclusion of other social media platforms. COSMOS was starting to broaden from Twitter to YouTube. So I was kind of helping to do that, but with Reddit—although, as soon as I graduated, Reddit shut down their public API.
My dissertation was focused on, again, trying to compare how people use language. This was a tricky kind of problem, though. In one instance, we were looking at how people talk about different religious identities. I got to take some of the things I was interested in from my bachelor’s degree in religious studies and apply that to this kind of work. With computational methods, how you frame the comparison is important; it’s really easy to do a comparison that just says, well, people with identity A use these words that are strong signals of having that identity. So if someone is talking about Christianity, they’re going to use words that we associate with Christianity. We can think of them as the technical language or jargon of Christianity. And that’s necessarily different from a discourse that’s about, say, Buddhism. Sure, people use these really distinguishing terms that signal membership in this identity, but let’s figure out how to suppress superficial distinctions between discourses in order to try and draw out deeper insight. How do they actually discuss things, or have a worldview? What is the deeper language game that people play in this discourse versus this other one? These were the things I studied.
Since leaving COSMOS, what roles/positions/jobs have you had? What is your current work?
Once you finish a PhD program, you basically have two options, which is one, find an academic job, such as a professor position, or do postgraduate work. Dr. Agarwal was kind enough to keep me around for another nine months as a postdoc in 2021. Then I started as an assistant professor at the University of Central Arkansas (UCA) in January of 2022. I’m in my sixth semester now at UCA, wrapping up my third full calendar year there. I teach our department’s artificial intelligence classes, one 3000 level class and a more focused machine learning 4000 level class, which are a mix of undergraduate and graduate students.
As you can imagine, it’s a double-edged sword, with all of the attention that AI gets at the moment. I’m almost intimidated, but also honored, to be trusted with the AI machine learning courses at UCA.
What positions did COSMOS and your classes at UALR best prepare you for?
So in terms of coursework, there was of course Dr. Agarwal’s class on social computing. Dr. Xu also taught a class on network science, and those had a big, big impact on me.
But aside from coursework, one thing that was, I think, really, really important was just getting exposed to so many different research papers—that I might not have sought out on my own—through the COSMOS Friday meetings, where everyone presented a paper. Those meetings were such a nice explorative approach to learning new ideas.
Everyone would pick different kinds of papers, and I would always pick these outlandish or ’out there’ papers. We’d have a really good discussion about them, and it motivated me to read papers and try to dig into papers that other people had presented that I had questions about. There’s a saying: one hour in the library is worth 1,000 on the bench (or in the lab)—that reading, you can just sit there and try to figure everything out on your own. But that’s silly because science is a social enterprise. It’s really important to just see what people are doing, and that breadth of exposure was great. And that’s what I loved about Cosmos. All kinds of topics that fit in that span from the computational to the human, psychological, social, psychological, all those things.
If you had to describe the most momentous event at COSMOS, what would it be?
I once presented at a conference that was the International Conference on Complex Systems. It was in Massachusetts that year. And I got to meet one of my intellectual heroes who gave one of the talks at that. That was great, because I only got to go to that because of COSMOS and the funding that was available to me through that.
Another significant time was at the conference Computational Humanities Research in 2020, the first year that that conference came into existence. I was the lead author on a paper where I felt like I wasn’t just playing it safe, where we were toying with this problem of comparing religious discourses and trying to compare them in a way that didn’t just reflect these superficial distinctions in their jargon, more or less. It’s really easy as a grad student to want to do ‘safe’ science—we take the data, we have the methods, we apply them, we report the results, we send the paper off. But this was one of the first times that I felt enough confidence to say, “Here’s kind of a new idea,”—where people haven’t done this specific thing before, and so wading out into the unknown. It’s a little bit scary doing that, especially as a grad student. But that paper got accepted, and the people who were there at the conference, who were in that session, were excited about it. Dr. Agarwal mentorship through all this was pivotal—the fact that he kind of just let me pursue these ‘weird’ questions that I was interested in, saying “Hey, you know, if you can get it published, then go for it!”
What advice would you have for current Cosmographers?
On the one hand, the safe approach to research that I mentioned, it serves a function in terms of how labs and institutes have to sustain themselves on the currency of publications.
But I would tell everyone that one thing they can do is research that is going to be directly useful to somebody that they don’t know—that someone’s going to read that paper and think, oh, wow, that actually helps me. A lot of times, it’s easy to get stuck doing these papers that don’t really push the envelope. They’re fine. There’s nothing wrong or untrue about them. But there’s a risky kind of paper to do, to push the envelope, that could be hard to get accepted depending on how much it diverges from what most people might expect. But I feel that, especially for PhD students, if you can do something that really has your stamp of originality, then that is something that you will carry with you even if you don’t end up in an academic career or anything like that. And especially because if you do, that’s such an important thing to do at least once. At least give it a try. Trying to do a paper that’s exciting and weird and different, at least once.
Also, read widely! That too.