All academic fields can benefit from AI applications. “When we don’t use AI, we are working with historical samples only,” explains Kara Kuebart, who is a Research Associate at the University of Bonn Department of History. “The volume of data is simply so enormous that it is literally impossible to work through everything taking a manual approach. Machine learning methods are crucial if we are to fully evaluate the available data.”
Kara Kuebart thus started in 2022 developing seminars and exercise classes on the subject so future historians will have acquired AI competency in the course of their studies. “The aim is to impart digital skills enabling them to create maps, digitally analyze texts and make use of computer data processing, for example,” she relates, “so students feel more comfortable utilizing code-driven AI programs."
"I had a lot to learn myself starting out, and began by experimenting around with the Python programming language.” Then she heard about the “AI for Everyone” seminar offered by the University of Bonn as part of the BnTrAInee project, the purpose of which is to bring our AI know-how existing within Computer Science out to the many different academic departments so they can develop needs-based teaching and courses.
The long road to AI expertise
“The intensive courses offered teach the fundamentals of using artificial intelligence methods,” explains Dr. Moritz Wolter, who co-designed the courses together with his colleague Dr. Elena Trunz. “The only prerequisite is completing a tutorial in Python.”
Students of the natural sciences generally have more affinity, as “they already gain familiarity with the mathematical foundations as part of their studies.” Herself a student of the humanities, Kara Kuebart rose to the occasion—and discovered that machine learning can actually be fun. While learning the mathematical basics, students taking the intensive courses are tasked with solving various programming problems. “One of the most popular class exercises is to train a language model exclusively using the works of Shakespeare,” she relates. “The texts then generated by the model do sound like Shakespeare ... though artistically they are not as sophisticated.”
“Moritz and Elena have structured the courses quite expertly, exploring the different types of AI, explaining how to build these models and providing an overview of potential applications,” says Kara Kuebart. “That makes it easier to figure out what kind of AI model to use for a given research inquiry.”
Now she is passing on the knowledge she has gained to her own students. “There is a very high level of interest. In a recent course we created maps using a programming library, for which everyone had to write codes. It worked out really well.” Kara Kuebart uses artificial intelligence methods in her own research, as she elucidates: “As an example, at our chair we use AI to search through old editions of the Kölnische Zeitung newspaper from the 19th and 20th centuries to identify political mood shifts in articles concerning the First World War. It is also used to analyze help wanted ads for housemaid positions to study the discourse at that time and aspects like the qualifications sought.”
Nearly 100 researchers have taken the courses thus far, and the “AI for Everyone” series is constantly being optimized. “We are focused on the needs of our student clientele,” says Dr. Wolter. “In time, more advanced courses will be offered to go more into depth on particular focuses.” Kara Kuebart herself is quite interested: “I will be definitely be taking further advanced courses when they are rolled out.”
The Digital Strategy of the University of Bonn
The University of Bonn’s Digital Strategy (www.digital.uni-bonn.de) sets out the measures and structures required for its digital transformation. The BNTrAinee project is an initiative forming part of the strategy’s Digital Skills set of objectives and aims to build AI expertise across discipline boundaries. Find out more about the Digital Strategy at https://www.digital.uni-bonn.de/en.