Get to know: Lin Ma

Lin Ma is an assistant professor in CSE who joined the University of Michigan in Fall 2023. His research concentrates on the intersection of database management systems and how machine learning techniques can be applied to automate and optimize database administration.

Ma’s research has been published widely in top conferences in the field, including ACM SIGMOD, VLDB, and more. Ma earned his PhD in computer science at Carnegie Mellon University in 2021, after which he worked on the Delta Lake/Lakehouse data platform at Databricks. 

We recently sat down with Ma to learn more about his interests as a researcher and professor.

What are the key research problems that motivate your work?

My work primarily focuses on the performance and usability of database management systems. Database systems have become essential in almost all aspects of society, ranging from leveraging data-science tools to make data-driven decisions to using large amounts of data to train artificial intelligence (AI) and machine learning (ML) models. However, as data volume and diversity increase, database management systems are also becoming more complex with a growing list of functionalities. Thus, optimizing, deploying, and maintaining the performance of database management systems has become increasingly challenging. 

What’s unique about your approach to tackling these problems?

My approach especially leverages AI and ML techniques to optimize the performance of database systems and make them easier to use. Recent advances in hardware and AI/ML algorithms have made it possible to collect abundant telemetrics about a database system’s performance and train ML models to predict this performance. This gives us the opportunity to use these models to capture the complex interactions of the system’s various functionalities that are difficult for humans to reason about. Therefore, the database system can use these models to intelligently optimize and maintain its performance.

How do you see your work impacting society at large?

Given the importance of database systems in various data-driven applications, their performance and ease of use are evidently important. Furthermore, I think the lessons we learn from applying AI/ML to databases can also help others develop intelligent software systems, especially for complex systems that are challenging to optimize manually. 

What are your future goals with regard to research?

Even though there have been significant academic interests and promising research results on applying AI/ML for databases in the past few years, the real-world applications of such techniques are still limited. Given my past research and industry experience, I think there are a few gaps between the research and practice of ML-enhanced database techniques that I want to address in the future, including the robustness of such techniques in volatile environments, the considerations of such techniques in a larger data pipeline, and the reuse of knowledge learned from different database systems.

What’s most important to you as a mentor to graduate students?

I think the most important aspect of a graduate student is probably their motivation to do research. There is always room to learn new things, but whether one has a genuine passion for research and the tenacity to face challenges in research may eventually determine their success in a research career.

What do you expect from the students you work with?

I think the growth of graduate students generally goes through a few different phases, so the expectations can vary over time. At a high level, in the beginning, the students may need to spend time picking up research skills and methodologies, including both techniques, writing, and speaking. As the students become more mature as researchers, they are expected to independently identify valuable and challenging research problems, plan out the research agenda, and execute the plan potentially with a team of more junior researchers.

When you’re not thinking about computer science, what else do you do?

When I have free time, I like to watch basketball games and movies.