# A note for prospective students.
This is a long-due post. As my gap year at Microsoft Research goes by quickly, I am starting my full-time position in the Department of Computer Science at Johns Hopkins University in Jan 2024. That being said, the most important part of my time will shift to working with students. In this note, I want to share some initial, immature thoughts on student advising and the future research landscape.
### On Student Advising
I have been lucky to have three great advisors during my time at UIUC, Prof. Dov Cohen, Prof. Karrie Karahalios, and Prof. Hari Sundaram. They shaped the way I approach research, work with people, and interact with the world. I hope I can pass on what I learned from their mentorships to my future students.
The overarching goal for my Ph.D. advising is to prepare my student, potentially you, to be an independent researcher after graduation.
To achieve that I think there are three aspects,
* **Nurture curiosity.** Motivation is the key. My curiosity about the surrounding reality keeps me doing what I am doing. Although obstacles, frustrations, and confusion are often in the space between you and the answer, curiosity keeps you moving ahead. We will spend time discussing provocative ideas, speculating the future, and critiquing the past. My goal is to create a safe lab environment to express ideas and ask questions. I hope your curiosity about the universe only grows after your graduation.
* **Build your research tastes.** There are millions of questions worth studying, but, unfortunately, we are facing constraints on things like knowledge, time, funding, etc. It is important to make wise choices to devote your time and effort. During my time with my advisors, the most important thing I learned (and still learning) is how to find topics that both you and your audience are excited about and devote your time to. A criterion for your graduation is your ability to independently evaluate ideas.
* **Be open-minded.** Validating ideas is the most exciting process in research. You need to build your research skillset to operationalize your thoughts. What kind of research skills to master? It depends, depends on the question you are interested in instead of what discipline you are in. This is the beauty of human-computer interaction and doing interdisciplinary research, where you can build up research skills from multiple disciplines. Therefore, you need to keep your eyes open to new ideas and novel research methods. You may not have time to learn them all but knowing what to learn is important.
* **Support others.** Your cohort, colleagues, friends, and families are the most valuable support you have. In my lab, supporting each other does not only mean collaborating on research projects, sharing technological expertise, and discussing interesting ideas but also building healthy relationships to survive the most turbulent waves. You need to learn to be a strong ally to provide support to whom you may not share many of the experiences they struggled with.
I am still learning towards them but I am fortunate enough to learn from others. Now, I think I am in a position to share my experience with you and help you navigate through the mist of knowing.
For a more concrete advising style, my starting point is grounded on the great [Ph.D. Syllabus](https://docs.google.com/document/d/11D3kHElzS2HQxTwPqcaTnU5HCJ8WGE5brTXI4KLf4dM/edit#) from Prof. Eric Gilbert.
### On Research Landscape
I see myself as a computer science researcher with a background in psychology, computer science, and statistics. So the questions I am most interested in resonate mostly with human-computer interaction communities. My research is motivated by the fundamental question of understanding humans at scale. This is a very big umbrella and there are many ways to approach this question. I would like to actively explore the following directions in the coming years,
* **AI4SocialScience.** As an individual, my contribution to understanding humans is quite limited. So, my goal is to empower scientists, who study human behavior, to operationalize intuitions and curiosities about how we think and behave. This stream of research aims to understand the role of AI in social science research beyond efficiency and scalability, e.g., how we could conduct robust and generalizable studies about human behavior.
* **Model Evaluation.** I see model evaluation as a crucial instrument to close the sociotechnical gap between what people want and what technology can offer. One focus of this direction is to understand what makes an evaluation method (e.g., benchmark, metrics, etc.) effective and in which context an evaluation method is effective. Another focus is to develop contextualized and robust evaluation methods that are grounded on an individual’s real needs. If you want to know more, here is a recent [position paper](https://arxiv.org/pdf/2306.03100v1.pdf).
* **Power Dynamics in Human-AI Interaction.** With respect to human behaviors, I am particularly interested in the relationship between power and information. As we are living in an information-rich world, the information we seek, consume, and disseminate could have profound real-world consequences. And power is intertwined with what information we have access to. I hope this line of research will shed light on how power manifests in our daily interaction with intelligent systems, their impact on our information decisions, and the individual’s agency and control of information access.
### Final Thoughts
If you resonate with this note and consider applying for grad school in the coming cycle, I am recruiting 1-2 Ph.D. students at the CS department of Johns Hopkins University. Boht of them are fully funded. If you are interested in applying, here is the [link](https://www.cs.jhu.edu/academic-programs/graduate-studies/phd-program/phd-program-admissions/). Regarding my department and Ph.D. application process, I highly recommend my colleague Prof. Mark Dredze’s great [introduction video](https://www.youtube.com/watch?v=YlwpIPFNXjo&t=730s).
I am looking forward to reading your application. Feel free to drop me any questions. And good luck with wherever you applied to!
Updated: 8/28/23