About the Machine Learning Center

The Center supports research teams representing a wide range of academic disciplines, assisting them in selecting and effectively applying computational tools, analysing data, and implementing machine learning methods in their research. Artur Kalinowski, Professor and Head of the Machine Learning Center, spoke with Patrycja Chuchała, MSc, about the Center’s activities, its role in advancing interdisciplinary research, ongoing projects, and plans for future development.

What inspired the establishment of the Machine Learning Center?

The establishment of the Machine Learning Center was one of the initiatives implemented under the “Excellence Initiative – Research University” (IDUB) programme. The idea emerged in response to the need to provide the academic community with access to specialised computational expertise and tools capable of significantly enhancing and streamlining the research process.

What needs within the academic community is the Machine Learning Center intended to address?

Modern computational tools can substantially improve the efficiency of research, but their effective use often requires specialised expertise and appropriate technical infrastructure. Not all research teams have access to such competencies. The Center therefore provides support from specialists with expertise in programming, data analysis, and machine learning, who assist researchers in the practical application of available technologies.

What distinguishes the Machine Learning Center at the University of Warsaw from other initiatives involving artificial intelligence?

A distinctive feature of the Center is its centralised model for organising and funding specialised support. Specialists employed by the Center can collaborate with multiple research teams, while their expertise and experience remain available for subsequent projects. This model eliminates the need for each research group to independently recruit a specialist for a specific task. Moreover, knowledge and expertise acquired through individual projects can subsequently be transferred and applied to new research initiatives.

How does the Center collaborate with research teams?

Research teams can approach the Center with specific problems requiring computational tools or data analysis methods. After assessing the team’s needs, the Center’s specialists work together with the researchers to define the scope of the project, select appropriate solutions, and develop the tools required to address the research problem.

What kind of support does the Center provide to researchers?

The support primarily involves data analysis and processing. In practice, this may include developing computer code that transforms input data into outputs required to address a specific research question. The Center also assists researchers in configuring appropriate computational environments and selecting suitable analytical tools.

Which academic disciplines does the Machine Learning Center support?

The Center is open to collaboration with researchers from all academic disciplines. To date, projects have been carried out in cooperation with, among others, the Faculty of Biology, the Center for Latin American Studies, and the Faculty of Archaeology. The diversity of these projects demonstrates the broad applicability of machine learning and data analysis methods across a wide range of research areas.

Has there been an increase in interest in machine learning within the humanities and social sciences?

Interest is very high. In many cases, the application of machine learning primarily involves automating processes that previously required extensive and time-consuming manual work. Advances in computational tools have made such tasks considerably easier and more efficient.

One example is the Center’s collaboration with the Center for Latin American Studies, which involved the analysis of scanned historical sources. The material comprised approximately one thousand pages, and the objective was to identify specific words and phrases within the documents. The application of appropriate computational tools significantly accelerated and streamlined the retrieval of relevant information from this extensive collection.

Do the projects make use of artificial intelligence models, including language models?

Yes. However, the Center’s primary area of focus at present is image analysis. Existing tools for analysing visual data are employed and subsequently configured and adapted to meet the specific requirements of individual research projects.

What skills should researchers have if they wish to collaborate with the Center?

Advanced computational skills are not required. The most important requirement is to clearly define the research problem and identify the data to be analysed. The Center can then help determine which technological solutions are most appropriate for addressing the problem.

It is important to recognise, however, that artificial intelligence is not a tool that can automatically solve every research problem. In machine learning applications, the availability of a sufficient quantity of high-quality data, including appropriately labelled data, is crucial. Preparing such datasets is often one of the most significant challenges and limitations encountered in these projects.

Does the Center provide training or other educational activities?

In the past, the Center organised short, one- or two-day training courses held twice a year. For organisational reasons, however, this format was subsequently discontinued.

The Center is currently considering the development of more targeted educational activities organised in cooperation with individual faculties and research units.

Such an approach would make it possible to first identify the specific needs of a given academic community and subsequently design training programmes tailored to the particular challenges, research interests, and requirements of its researchers.