Information studies could be at the forefront of the AI transformation – Paavo Arvola examines how AI is changing the way we seek information

Paavo Arvola is a University Lecturer in digital humanities. Since 2021, he has also served as Head of the Degree Education in Multidisciplinary Communication Studies. Of all the fields represented within this programme, information studies is the only one that Arvola has formally studied, although his academic journey took him through four major subjects before graduation.  

When Arvola applied to university, he did not have a specific career path in mind but felt that university was the right choice. He initially chose mathematics and statistics as his major subject. He later applied to study philosophy before eventually transferring to computer science, the field in which he completed both his bachelor’s and master’s degrees. 

“What fascinates me about computer science is the opportunity to learn by doing and see the final results. I originally planned to graduate in computer science and pursue a career in industry or start my own company, but I came to realise that academia, too, offers opportunities to be bold and pursue the truth.” 

While working towards his doctoral degree, Arvola switched his focus to information studies and interactive media. His dissertation, titled The Role of Context in Matching and Evaluation of XML Information Retrieval, explored focused information retrieval. 

“I did not start out as a digital native who was born with a computer mouse in hand. Even so, my studies in computer science gave me a strong foundation that I continue to draw on in both my research and teaching. They also taught me disciplined thinking.” 

Many roads lead to information studies  

Arvola is grateful that his studies gave him the opportunity to try different things. 

“This kind of flexibility is probably not even possible today because of the time limits placed on degree completion, and it may even have been discouraged when I was a student. However, my own experience at university continues to shape the way I think about the planning and management of education: in some degree programmes, the introductory studies could be broader in scope, allowing students to narrow down their focus at a later stage.” 

According to Arvola, students enter information studies from a wide range of backgrounds. Some are changing careers and have previously worked in entirely different fields, but their prior experience should be seen as an asset. As a discipline, information studies benefits from students with diverse interests and Arvola has yet to encounter a field that could not be combined with information studies. For example, the information behaviours of doctors, engineers, historians and journalists are not alike, and information studies needs this diversity of expertise, interests and perspectives. 

“Information studies is concerned with how people seek and use information. The discipline can be described as democratic because it examines information from the perspectives of the recipient, the seeker and the user, instead of focusing solely on its production and dissemination. An effective democracy depends on an educated population, and the value of broad knowledge and specialist expertise has not diminished. On the contrary, education and a willingness to read remain as important as ever.” 

Practices changes, concepts endure 

From the outset, Arvola’s areas of expertise within the broader field of information studies have included information retrieval and information interaction. His current research and teaching focus on the role of information retrieval in AI systems.  

“There are, of course, many serious concerns surrounding AI. Yet from an academic perspective, it is exciting and intellectually stimulating to be involved in this period of transformation. I would be happy to see information studies play a central role in these developments. As large language models are changing the way people seek, formulate and evaluate information, many of the key questions raised by the AI transformation are the very same questions that have occupied information studies for a long time: What is relevant information? How do we find it? How do people express their information needs? How do we evaluate the outcome?” 

One of Arvola’s current areas of interest in both research and teaching is information interaction, where users guide AI systems through prompts. As part of their coursework, students are tasked with developing aprompt that instructs an AI system to transform the user's information need into a search query that reflects the user’s information needs. This query is then entered into a search engine, and its effectiveness is evaluated based on the relevance of the retrieved documents and how highly the relevant documents are ranked in the search results. Through this process, students investigate the effectiveness of prompts empirically while also learning more about how prompts are formulated and how AI systems operate. 

Photo: Karu Films

Arvola aims to integrate his research with the development of teaching. He believes that teaching methods should also be developed based on research evidence in response to the rapid changes brought about by AI. Arvola has presented the methods he has developed at scientific conferences and other academic forums. He recently presented the article “From Prompts to Queries: Teaching and Measuring Prompt Engineering as Query Construction”, which he co-authored with Tuulikki Alamettälä, at the SEFI 2026 Annual Conference organised by the European Society for Engineering Education (SEFI) in Prague. 

“The conference focuses on engineering education. My teaching is not engineering as such, but the methods are similar: designing experiments and measuring the effectiveness of prompts.” 

Although AI is reshaping information studies, along with many other fields, new innovations continue to be built on the foundations of earlier ones. For example, in the coursework described above, the effectiveness of prompts is assessed using metrics developed by information researchers in Tampere in 2002 to measure cumulative gain. 

There are also many established concepts that remain relevant, one of them being relevance itself, which is a key criterion for evaluating the outputs of AI systems. At the same time, the age of AI raises questions about the applicability of existing concepts. 

“We can ask, for example, to what extent ChatGPT is an information retrieval system. When it generates a response, should that response be considered a document? We still need the established concepts, but we must also consider what they mean in this new environment."