Finding structure in multi-modal food data with visual analytics
PhD opportunity at KU Leuven: Finding structure in multi-modal food data with visual analytics.
The Augmented Intelligence for Data Analytics (AIDA) research group in the Department of Biosystems at KU Leuven is recruiting a PhD candidate to work in a highly interdisciplinary project on healthy and sustainable food. The position is led by Prof. Jan Aerts and is embedded in the HSFood4ALL (Healthy and Sustainable Food for All) project, which combines expertise in nutrition, consumer behaviour, sustainability assessment, governance, data science, and participatory research.
This PhD focuses on the analytical and visual backbone of the project. The research problem is to uncover structure in complex, multi-modal food data by joining information across healthy food choices, environmental sustainability, and voluntary sustainability standards. The work is exploratory and open-ended: using unsupervised machine learning, topological data analysis, and interactive visualisation to surface patterns that are not visible in siloed analyses.
The project has three connected research themes. First, discovery: identifying cross-facet structure such as nutrition-environment trade-offs, consumer segments, and dependencies across datasets. Second, integration: designing data representations and schemas that allow heterogeneous sources to “speak to each other” for downstream analysis, including relational, document, and graph-based approaches. Third, translation: developing effective front-of-package visual formats that remain faithful to the data while being understandable to consumers and policymakers.
The successful candidate will build on AIDA’s methodological line, including tools such as minimum spanning tree-based topological networks, flare-sensitive clustering, and PLSCAN, and extend these methods in the context of food systems. The work will be collaborative, with domain experts providing questions and ground truth, and with research outputs aimed at visual-analytics and data-science venues such as EuroVis and IEEE VIS.
Eligibility highlights include an EU Master’s degree in Data Science, Bioscience Engineering, Computer Science, or a related field; familiarity with clustering and dimensionality reduction; strong Python skills; SQL knowledge; and an interest in visual design beyond standard plotting. Experience with NoSQL, document databases, graph databases, or topological data analysis is advantageous. Dutch is a plus.
The position is offered as a one-year appointment, with the possibility of extension to four years after positive evaluation. The successful candidate may enroll in the Arenberg Doctoral School. KU Leuven also highlights opportunities for advanced training, international conferences, collaboration with academic, industry, and policy stakeholders, and a supportive international environment.
Applications must be submitted via the KU Leuven online application tool by 2026-08-14. Screening is rolling, so early submission is strongly encouraged. Required documents include a short motivation letter, CV, academic transcripts, and contact details for at least two references.
