Scientists are working with increasingly complex datasets. This creates many new opportunities for research, but also introduces challenges in, for example, interpreting the data and sharing it with the scientific community. How can researchers make better use of their data and share it more effectively? This is one of the questions being explored by Lise Stork, Assistant Professor at the Informatics Institute and MacGillavry Fellow.

Stork's research lies at the intersection of three research areas. Stork: 'First, FAIR data, which stands for findability, accessibility, interoperability, and reusability. Then, hybrid intelligence, which focuses on how we create AI systems that amplify human capabilities rather than taking them over. And finally, e-Science, which focuses on supporting scientists in their work.'

Detecting outliers

Besides enormous datasets, many researchers also work with various types of data, such as images, texts, and tables. Integrating this data to extract insights is often a significant challenge. A PhD candidate in Stork's group is working on ways to extract more insights from biodiversity data.

This dataset consists of various types of animal data, such as images and tables with locations where they were found. Stork: 'These tables often contain small errors. My PhD candidate is looking at how he can use different types of data to identify outliers in the tables and explain why they are outliers. Then, using their domain knowledge, the researchers can determine whether it's an error in the data or new information.'

Help from AI

When scientists share their datasets, it's important to also describe the context, such as how the experiments were conducted exactly. According to Stork, it's often difficult for researchers to determine which information is important. Stork: 'A PhD candidate in my group will develop workflows for this, where scientists can collaborate with an AI model to better describe the context of a dataset, making reuse easier.'

AI can also act as a facilitator between different researchers. According to Stork, this would make it easier to communicate with scientists from other disciplines. 'I notice that even in project meetings within computer science, everyone works with slightly different frameworks and formalizations, which complicates communication. With the help of AI, you could more easily describe the data to people both within and outside your own discipline.'

Developing a focus

In 2024, Stork started as an Assistant Professor at the UvA through the MacGillavry Fellowship. She has thoroughly enjoyed the transition from Postdoc to Assistant Professor. 'I think the courses I teach align perfectly with my research interests. I also really enjoyed shifting my focus to writing proposals, and I'm starting to find my way in supervising PhD students.'

In the future, Stork believes it's important to develop a clear focus. 'Building on the students I'm already supervising within my research theme, my biggest goal is to   establish a research line at the intersection of hybrid intelligence, FAIR data, and e-Science.'