Data Shack

Within the academic research activities of the Gianfranco Ferré Research Centre, the collaboration with the Harvard John A. Paulson School of Engineering and Applied Sciences (SEAS) is developed in synergy with the Data Science Lab and the Fashion in Process Lab of Politecnico di Milano through the Data Shack project. The initiative involves students and faculty from the Harvard SEAS Master’s programmes in Data Science and in Computational Science and Engineering, alongside participants from the Politecnico di Milano Master’s degrees in Computer Science and Engineering, Communication Design, and Design for the Fashion System.

The Data Shack project presents students with challenges based on real-world cases, requiring the application and development of skills including data management, machine learning, data analysis and visualisation, statistics, mathematics, and user experience design. Students are guided through the development of methodologies, software, visualisations, and high-performance components, leading to testing, solution finalisation, and the drafting of scientific papers.

During its first edition, the project developed with the Gianfranco Ferré Research Centre drew on the extensive digital archive to create an innovative vector database model capable of integrating machine learning techniques for the development of cultural ontologies. The initiative combines expertise in fashion design, data science, and human–computer interaction to experiment with new methods of semantic classification and interaction with archival content. It also explores co-creation processes within the phases of creative development, with the aim of connecting fashion’s cultural heritage to innovative design approaches.

In its second edition, the project further investigates the use of generative artificial intelligence tools – including multimodal models, LLMs, and related technologies – with the objective of developing pipelines capable of improving the performance of AI tools applied to fashion, particularly chatbots, LLMs specialised through fine-tuning, and other advanced solutions.

The outcomes of the project open up new perspectives for the creative use of archives. Dialogic interfaces based on generative AI enable more immersive and personalised exploratory experiences, benefiting research, education, and the enhancement of fashion’s cultural heritage.