We have successfully held our first retreat

With NERDS hitting 7 years old and growing over 30 people, it has been long overdue to hold our first retreat. We did so last week at the AI Pioneer Center, in the old astronomic observatory inside Copenhagen’s beautiful botanic garden (plus awesome dinner at Food Club Nørrebro), organized by Jonas, Jan, and Toine.

The retreat was a wonderful event that deepened our social ties, where we learned much more about each other, and discussed what works well or what does not work so well at NERDS.

On all levels, from faculties to long-term NERDS and visitors, we identified issues we want to improve, including diversity and hiring, more internal exchange, website updates, or more formalized tasks with ownership (and PhD duty credits). It was great to see though that we do not have any serious social issues – to the contrary, the retreat was a confirmation of how well we all get along, and how nice a research group can be. ❤️

In the coming weeks we are going to get to work to implement the short-term-implementable issues, keep pushing for improving our long-term issues to make NERDS an even better place, and definitely aim to make the NERDS retreat a recurring experience!

 

Ariel Avanzi has joined NERDS

Ariel joins us as a new Research Assistant. With his background in the physics of complex systems, he will work with Jonas L. Juul on quantifying how networks change.

One important application of Ariel’s work could be in the shipping industry, and the project is funded by two maritime foundations: Orient’s Fund and the Danish Maritime Fund.

We are thrilled to have you on board, Ariel.
Ahoy!

New NERDS publication on transport network growth

The trade-off between directness and coverage in transport network growth, by C. Sebastiao, A. Vybornova, A.R. Vierø, L.M. Aiello & M. Szell, published in Applied Network Science

We systematically study the growth of connected planar networks, quantifying functionality of the growing network structure. We compare random growth with various greedy and human-designed, manual growth strategies. We evaluate our results via the fundamental performance metrics of directness and coverage, finding non-trivial trade-offs between them. Manual strategies fare better than greedy strategies on both metrics, while random strategies perform worst and are unlikely to be Pareto efficient. Centrality-based greedy strategies tend to perform best for directness but are worse than random strategies for coverage, while coverage-based greedy strategies can achieve maximum global coverage as fast as possible but perform as poorly for directness as random strategies. Directness-based greedy strategies get stuck in local optimum traps. These results hold for a number of stylized urban transport network topologies. Our insights are crucial for applications where the order in which links are added to a spatial network is important, such as in urban or regional transport network design problems.

Two New NERDS Papers: Politician Campaigns; and Money Laundering

We have two new publications out!

  1. Disconnect between the public face and the voting behavior of political representatives by Christian Ivert Andersen and Michele Coscia, published in the journal Applied Network Science.

    One of representative democracy’s tenets is that a political candidate runs on a specific platform, which is information the electorate uses to determine whether to vote for them or not. If this promise is to be maintained, it is fundamental that the public face candidates present corresponds to their actions in parliament once elected. Such a promise has been put in question both by scholars, but also by the electorate. In different countries at different times, the people have expressed various degrees of dissatisfaction with democracy: often the feeling is that representatives put their own interests—or the interest of a powerful minority—before the ones of their constituencies. In this paper, we propose a network-based quantitative investigation of this disconnect between the public face and the voting behavior of elected representatives. By using data from Denmark, we can place politicians in two different spaces, determined by their electoral campaign promises on the one hand, and on the other hand by the votes they cast in parliament. We find that our technique makes it possible both to find clear, expected, and consistent left-right divides between the political parties; as well as a larger-than-expected disconnect between the public face and the voting behavior. Our preliminary results indicate that the aggregate voting behavior in parliament of politicians does not match with how they present themselves to the public on the salient issues discussed during the election campaign.
  2. Evaluating fraud detection algorithms in a decentralized scenario by Ada M Gige, Lasse Buschmann Alsbirk, Michele Coscia, published in the journal Royal Society Open Science.

    Financial fraud is an umbrella term including a vast number of illegal activities. These activities involve a significant fraction of the global economy. Traditional investigation techniques are labour-intensive and cannot scale to match the size of the issue. Machine learning has provided effective tools which deliver high accuracy in identifying transactions that could be involved in fraudulent activities. In this paper, we point out that the state-of-the-art in financial fraud detection has been applied to the unrealistic scenario of an omniscient centralized global authority which has access to all bank transactions globally. We propose a more realistic evaluation scenario, one made of two steps: first, the bank flags its own transactions using exclusively information it possesses; then only flagged transactions from all banks are analysed by the governmental authority for potential prosecution. We find that, in such a realistic scenario, the effectiveness of the state-of-the-art method for financial fraud detection decreases. Moreover, we show that in this decentralized scenario, it pays off to use simpler methods than the state-of-the-art, depending on the specific objective function the system wants to ensure.

New NERDS publication on candidate recommendation

Analyzing the Effects of a Human-in-the-Loop Candidate Recommendation Algorithm at Jobindex, by Mesut Kaya & Toine Bogers, published in ACM Transactions on Recommender Systems

Recruiting is the process of assessing relevant candidates for an open position based on their education, work experience, and knowledge, skills, and abilities. As part of a collaborative project between academia and industry, we developed an automated candidate recommendation system to support recruiters in this time-consuming task of matching CVs to job postings. We chronicle the development and deployment of a candidate recommender system at Jobindex, expanding our focus from a pure machine learning problem to a holistic overview of the development and deployment of such a system in a real-world setting. After extensive offline and online experimentation, we integrated our algorithm into the recruiters’ everyday workflow. Our second contribution is a detailed analysis of how these recruiters have adopted this candidate recommender system, and which factors influence their engagement with the slate of suggested candidates. In this article, we present the results of 17 months of data (corresponding to 41,390 jobs) and show how engaging with these recommendations has impacted the recruiters’ work, and which factors influence their task success. While adoption of the new system was initially hindered by deeply-ingrained habits and a lack of trust in AI, over time the combination of human and automated recruitment shows considerable promise across a variety of job and recruiter characteristics.

Two new NERDS papers: data storytelling visualization, LLMs for complex information needs

We have two new publications out!

  1. The Influence of the Communication Medium on Data Storytelling, by Tamara Nagel and Toine Bogers published in CHIIR ’26: Proceedings of the 2026 ACM Conference on Human Information Interaction and Retrieval 2026

    Despite increasing interest in data storytelling, it remains unclear how the choice of communication medium shapes its effectiveness, particularly for audiences with varying levels of data literacy. This paper reports on a controlled, longitudinal experiment comparing verbal to written storytelling alongside a baseline data visualization condition. Each condition employed simple graphs and an author-driven narrative to examine their effects on recall and attitude change. Results showed mixed results of data storytelling: while storytelling did not improve recall, verbal storytelling and no storytelling facilitated long-term attitude change, whereas written storytelling did not. Higher data literacy supported long-term recall but was associated with smaller immediate attitude shifts, an effect that diminished over time. These findings challenge assumptions about the universal advantages of narrative-based communication, demonstrating that medium, topic familiarity, and audience characteristics jointly determine outcomes. The study contributes empirical evidence to the field and calls for further research into how narrative structures and visualization complexity affect the effectiveness of data storytelling.
  2. Tip-of-the-Tongue Search in the Wild: Analyzing Human and LLM Performance and Success Factors on Complex Search Requests, by Toine Bogers, Maria Gäde, Mark Hall, Marijn Koolen, Vivien Petras, and Mette Skov, published in CHIIR ’26: Proceedings of the 2026 ACM Conference on Human Information Interaction and Retrieval 2026Users often turn to online forums when searching for known books, movies, or games that they cannot identify through conventional search engines. These “tip-of-the tongue” requests present a unique challenge, appearing highly variable in formulation, context, and specificity. So far, these could mostly only be solved by other humans answering in forums. Generative AI is believed to help solve these specific questions. In this work, we manually annotated 150 requests each for books, games, and movies in the casual leisure domain to study the differences between solved and unsolved requests and identify factors that influence their difficulty. We compare human responses in forum threads with the performance of a Large Language Model (LLM) under similar conditions. Specifically, we investigate how the formulation of requests affects human and LLM success; how item properties impact LLM retrieval; how interaction and feedback within a thread shape human and LLM performance; and whether increasing the information provided to an LLM improves its chances of solving the request. Our findings offer new insights into what makes these known-item search problems easier or harder to solve. This study contributes to a better understanding of complex search behavior and the role of LLMs in helping with difficult casual-leisure information needs.

Jonas Juul wins two grants for maritime-focused network science research

Orient’s Foundation and the Danish Maritime Foundation have decided to fund Jonas Juul’s research with a total amount of DKK 735.000 (€100,000.)

In the 18-month project, Jonas Juul will study the dynamics of growing and changing networks, developing new statistical tools to quantify how networks grow. The project will be conducted in coordination with the major Danish shipping company NORDEN.

We are excited that the maritime industry and maritime foundations have decided to board our network science journey.

Congratulations to you Jonas. We wish you fair winds and smooth sailing for the work ahead.

An AI-generated network of ships! (Created with Microsoft Copilot)

An AI-generated network of ships! (Created with Microsoft Copilot)

PhD and Postdoc call: Using data science to improve epidemic preparedness

We are looking for PhDs and Postdocs to join Jonas Juul improving inference, forecasting, and mitigation in future pandemics.

The project (funded by the Novo Nordisk Foundation) will improve epidemic preparedness in several ways, for example:

  • By improving the statistical foundation for the projection of epidemic case numbers.
  • By quantifying the effectiveness and efficiency of mitigation strategies analytically and numerically, we will get a better understanding of how mitigation strategies should adapt as an epidemic develops.
  • By studying plausible infection patterns in combination with Danish register data to understand how the COVID-19 epidemic propagated through Danish social networks.

The positions come with attractive opportunities for conference participation, travel, etc., and Copenhagen offers a world-class research community in network science, epidemiology, and data science.

Interested in large-scale data analysis, applied statistics, and/or analytical methods for studying epidemics in networks? Read more here:

and feel free to reach out to Jonas Juul (jjuu@itu.dk) if you have any questions.

Arianna Pera defends her PhD: The Language of Collective Action in the Social Web

Big congratulations to Arianna for successfully defending her PhD thesis on the role of language in fostering grassroots collective action. The Committee was composed by Luca Rossi, Fabiana Zollo, and Sune Lehmann. Her very proud supervisor is Luca Aiello.

Collage with three pictures 1) Arianna presenting her work, 2) Arianna posing with members of the Committee (Luca Rossi, Sune Lehmann) and with her advisor (Luca Aiello) 3) Arianna and many NERDS nighttime celegration

During her three years with NERDS, Arianna has published 7 paper in the areas of mobilization framing, computational narratives, and applied NLP. She has become a very active member of the international Computational Social Science community, which will benefit from her work for many years to come.

On March 1st she will start a research position at SODAS with Clara Vandeweerdt.

In bocca al lupo, Arianna!

Luigi defends his PhD: Polarisation in an Evolving Social Media Landscape

Big congratulations to Dr. Luigi Arminio. Yesterday he defended his thesis on Polarisation in an Evolving Social Media Landscape: Multimodal Insights from the Climate Debate.

Supervisor: Luca Rossi
Committee: Luca Aiello, Cornelius Puschmann, Cindy Shen

We are extremely happy and proud of Luigi’s research dealing with the crucial issue of online polarisation and producing both substantial contribution as well as methodological advancements.

Luigi’s work is built on 6 papers and investigated both textual and visual data. On the textual side he showed how opposite groups in a polarised debate are recognisable by specific socio-linguistic feature that we can detect computationally. On the visual side he contributed to the development of a research pipeline that leverages Visual Large Language Models to produce semantic clustering of visual content. This was then used to study the ongoing visual narratives surrounding the climate debate.

On March 1st Luigi will start a research position at Roskilde University in the context of the Algorithms, Data & Democracy project but we know that our collaboration isn’t over.