Featured AI2er: Bailey Kuehl
Bailey Kuehl is a Data Science Analyst on the Semantic Scholar team at AI2.
Tell us about your role on the Semantic Scholar team at AI2.
As a Data Science Analyst, I am responsible for communicating with the Semantic Scholar Research team regarding annotation projects and crowdsourcing needs. Along with offering expert annotations and data insights, I help ensure a smooth flow from the initial research idea to data collection and analysis. My day-to-day might include annotating a dataset, iterating on a project design with Semantic Scholar Research, and managing contract workers on Upwork to gather annotations.
What put you on the path to your current role?
For my undergraduate education, I pursued a degree in Materials Science Engineering and a double major in Biochemistry at the University of Wisconsin-Madison. My coursework, research, and internships all focused primarily on 3D printing of stainless steel. The project that introduced me to data analysis was a machine learning model that one of my labmates was using to predict the properties of 3D printed material based on the relationship among various input parameters in the printer. I then became intrigued by data and machine learning research, so I was delighted to find the Data Science Analyst position on Semantic Scholar. I was ready to make the transition to a more data-focused role, and I was especially excited that this particular role emphasized projects related to biomedicine, allowing me to utilize my background in biochemistry.
Give us a few highlights from your recent work on Semantic Scholar.
During the last 15 months, I have been fortunate enough to contribute to a wide variety of projects at Semantic Scholar, including SciA11y, TLDRs, VILA Structures for Scientific Text Classification, and more. Many of my contributions involved reading and annotating scientific publications in order to construct a dataset, such as the MS² dataset, which will be featured at EMNLP 2021 by Jay DeYoung. Other contributions included creating training data and evaluating the quality of the model outputs after training, such as my work on Dan Lahav’s novel search engine for discovering scientific challenges and directions. One of my favorite projects has been working on a follow-up to David Wadden’s SciFact work, which included assessing the validity of scientific claims based on corresponding evidence found by the proposed model.
What are you looking forward to with your work in the coming months?
I am extremely eager to continue working on the next iteration of the SciFact project. We have some exciting expansions to the dataset, and some interesting insights to analyze and share. I’m also looking forward to seeing the continued impact of Paper to HTML developed by Lucy Lu Wang and company. I’m hopeful that we will see projects in the future at Semantic Scholar inspired by this incredible work surrounding accessibility.
What is one piece of advice you’d give an aspiring Data Analyst?
Don’t be afraid to work outside your comfort zone! Expanding to tasks outside of your job description and challenging your current skill set will help you grow professionally and personally. It can be intimidating to take on a new role, or change your career path entirely. Hopefully, you will have a strong and encouraging support system to take those steps like I do at Semantic Scholar.
What are some of your favorite hobbies outside of work?
One of my favorite things about living in the Pacific Northwest is the scenery! I love trail running, exploring the mountains, and searching for wildflowers. When the weather keeps me from being outside, I enjoy baking (and eating), and spending time with my 2 year old cat, Monona.
The Semantic Scholar team at AI2 is hiring! Apply for an open position.