COSMOS, in this edition of Cosmographer Corner, highlights the work of a former University of Arkansas at Little Rock graduate and Cosmographer Dr. Tuja Khaund. Dr. Khaund—who is now a data scientist for Walmart—started her graduate education as a COSMOS research assistant. She received her BS in Computer Science, MS in Information Science, and Ph.D in Computer and Information Sciences from UA Little Rock in 2015, 2017, and 2021, respectively. During that time, Dr. Khaund worked at COSMOS as a graduate assistant. We interviewed Dr. Khaund on where her career is now and what her work at COSMOS entailed, with her responses below.

How did COSMOS fit into your university education career? How did you come across COSMOS and what were you studying when you joined COSMOS?

I came to the US in 2013 as an undergraduate exchange student, and Dr. Agarwal was one of my professors during my undergraduate days. So I’ve known him for a while. When I started my master’s, I reached out to him, and he offered me a project to work on, which later turned into a full-time research assistantship that continued until I graduated with my PhD. 

Specifically, it was a project funded by DARPA, where we were trying to understand social bots and their discourse and how they lead to disinformation and misinformation campaigns on social media.

I did Computer Science for my bachelor’s and then switched to Information Science for my master’s, but my PhD program was a combination of both. When I was working with Dr. Agarwal, he advised me to take more information science courses to better align with the focus of our projects, so that was part of the reason for the switch. 

What were some of the other projects you did during your time at COSMOS? 

From the beginning, I focused on inorganic accounts and research on which social media platforms have these accounts. We looked at it very extensively. We started with Twitter but also worked with blogs for disinformation studies, and I collaborated with my fellow Cosmographers on YouTube. So it was pretty diverse. We focused on different platforms or sources from where the information was coming, including blogs—that is, basically looking at different disinformation across different platforms, using graphs and network science theory.

I also worked in social cyber forensics regarding disinformation—trying to find the digital footprints of blogs. So you look at different IP addresses and try to tie them back to different kinds of other blogs or other accounts that are connected, to discover the initial set of blogs or sources spreading disinformation and fake news. For example, we looked at the different layers of a blog, essentially backtracking the digital footprint or the IP location, and this can be represented as a hierarchical graph, which shows nonpublic or hidden connections between blogs and users.

Since leaving COSMOS, what positions have you had? What is your current work? What positions did COSMOS and your classes at UALR best prepare you for?

I was hired full-time right after my graduation as, after leaving COSMOS, I joined Walmart  Global Tech as a data scientist, where I leverage AI and machine learning to build pipelines for various other businesses within Walmart.

As a data scientist at Walmart, I build end-to-end applications and help design pipelines for in-house products. This involves all kinds of work—getting the data, processing it, building the model, and presenting the results to the business partners through a custom dashboard. One popular tool between different organizations and teams, that I use, is Power BI (Business Intelligence), which allows interactive visualization of various connected metrics, charts, and data. I have also used Dash, a Python library to build custom dashboards for business use cases as it gives me full control of every aspect of my visualizations.

Similarly, dashboards can identify areas to focus on. Let’s say a store manager comes to me and says, “I want to see how much foot traffic my store is getting,” or “I want to see what time of day I have the heaviest traffic”. These questions mean finding the right kind of data to look at, and then presenting it for the manager in a form like a chart or graph. Dashboards are used as a form of storytelling— displaying visuals that’s easy for them to understand and make informed decisions. A few examples include store managers deciding what products to stock, where more staff are needed, what resources are required, and so on.

When I first joined Walmart, I also worked on a project that focused on natural language processing (NLP) and natural language analysis. At UA Little Rock, there were classes offered relating to NLP, such as the basics and baseline concepts of natural language, and those kinds of classes helped me understand how to address that kind of problem, how to build topics. Just in general, both data science and computer science classes prepared me for solving these kinds of problems; knowledge and experience from those classes helped a lot, such as learning how to deal with databases, how to prep data, and how to analyze or process it.

Additionally, in my work at COSMOS, we had many teams working on different tasks, and that also helped prepare me for work at Walmart, as the work there also involves working among different departments and teams.

If you had to describe the most momentous event at COSMOS, what would it be?

One of the things that I have cherished the most from my journey with COSMOS was my first attendance at SBP BRiMS (the International Conference on Social Computing, Behavioral-Cultural Modeling & Prediction and Behavior Representation in Modeling and Simulation). I think it’s a moment that’s pretty much shared between a lot of my Cosmographers, as it’s a core conference for social computing.

I presented my published work there and received a lot of meaningful feedback, and in general, the whole experience of going to my first conference was really just fun. I got to meet many subject matter experts in that field of work that I was doing research on, and I got to network with fellow researchers who were all either on the same level or maybe a bit senior to us—so I received many helpful tips. Sometimes it gets overwhelming when you start a PhD, because you don’t know where to start; there’s so much literature out there. You have to drill down to find a problem and figure out how to solve that—and my first conference was definitely one of those moments that helped with that issue.

What advice would you have for current Cosmographers?

I think for new Cosmographers, my advice is something I mention whenever I talk to students who are trying to pursue higher education: first you have to find a topic of interest and start looking at different conferences in that area. The COSMOS website Wiki has a curated list, but a great start is either SBP-BRiMS (the International Conference on Social Computing, Behavioral-Cultural Modeling & Prediction and Behavior Representation in Modeling and Simulation) or ASONAM (the International Conference on Advances in Social Networks Analysis and Mining). Once you have a good understanding of your work, and once you have delved deeper into the research, my advice would be to then go for journal publications if possible; having a publication is good for your educational portfolio, especially if you’d like to pursue a PhD or work as a postdoc. 

For industry roles, I suggest you pick a job family and start learning skills related to it. For example, if you want to take a data science role, I suggest you learn as much as possible and cover all the topics that fit under the data science lifecycle— including machine learning and statistics. Just in general, you don’t have to be an expert per se, but if you can become more knowledgeable in your area by picking up projects and tasks, then that helps the job search, especially if later down the line you have to learn and take up a new problem on-the-job.

And there’s always a lot of resources online for self-based learning. I think everyone should not rely just on school courses but also do as much as possible to pursue self-based learning. Continuously learning new skills that meet the industry standard means your prospects are that much better.