AKAndreas Kaltenbrunner

Andreas Kaltenbrunner

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Andreas Kaltenbrunner has a total of 116 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Rewarding Engagement and Personalization in Popularity-Based Rankings Amplifies Extremism and Polarization

    by Jacopo D'IGnazi, Emma Fraxanet, Andreas Kaltenbrunner, Gaël Le Mens, Fabrizio Germano, Vicenç Gómez

    Despite extensive research, the mechanisms through which online platforms shape extremism and polarization remain poorly understood, largely because observational settings make it difficult to identify causal effects. We propose a framework that analyses a specific mechanism: the feedback loop between popularity-based rankings and user behavior (e.g., partisan sorting and attention dynamics), and its consequences for content prominence. Our framework formalizes this feedback loop using a dynamical, popularity-based ranking model with a small set of interpretable parameters, and evaluates it through simulations and interactive experiments with hundreds of human participants. When trained on real-world data, the model reproduces well-established interaction patterns with ranked content: (1) users exhibit position bias, (2) prefer like-minded content, and (3) more extreme users engage more actively. Using this framework, we show that when platforms reward active engagement and implement personalized rankings, users are systematically driven toward increasingly extremist and polarized news consumption across a wide range of parameter settings.

  • Can large language models help in the assessment of people with pain?

    by Jacopo Amidei, Andreas Kaltenbrunner, Jose Gregorio Ferreira de Sa, Ruben Nieto

    1. Chronic pain (CP) represents a major global health challenge, with studies reporting prevalences that can be higher than 30% when including low and middle-income countries [1], making it one of ...

  • Open Access

    Are traditional hazard-based weather warnings anticipating local impact?

    by Jordi Morales, Xavier Llort, Andreas Kaltenbrunner, Agata Lapedriza

  • Open Access

    Using ML for the prediction of flood-related emergency calls

    by Jordi Morales Casas, Agata Lapedriza, Andreas Kaltenbrunner, Xavier Llort

    As weather-related disasters become more frequent and severe, there is a growing global push toward impact-based early warning systems, exemplified by initiatives such as EW4All. This transition positions machine learning (ML) and artificial intelligence (AI) as powerful tools for integrating meteorological hazard data with information on vulnerability and exposure into data-driven forecasting systems. In this work, we explore the use of 112 emergency calls as high-resolution impact proxies for an ML-based prediction problem. Specifically, we develop a model that combines rainfall-related weather data and static vulnerability-exposure layers to predict, at a municipal and hourly resolution, whether flood-related impacts will occur in the next hour. This study spans a period of over six years (October 2018 to February 2025) in Catalonia, northeastern Spain.To address the severe temporal class imbalance and uncertainty characteristics of emergency calls data, we define a custom walk-forward evaluation scheme that ensures the same number of positive samples across comparable time periods. We then distribute municipalities into three distinct population density groups (low, medium, and high) and train one model for each one. This stratification enables us to evaluate performance across diverse population dynamics and varying data availability. The resulting models are compared against operational methodologies, such as climatology-based weather warnings issued by meteorological agencies. Our results show that the ML approach represents a substantial improvement in two of the three groups. The model for the lowest-density group, however, struggles due to a substantial lack of impact data, highlighting a key roadblock for data-driven algorithm development in sparsely populated regions.To gain a more complete understanding and improve model trust and explainability, we perform a series of experiments: a feature importance analysis using SHAP (SHapley Additive exPlanations), ablation studies over different feature groups, and training models on individual feature sets. From these results, we can ascertain how the combination of varied data sources (such as weather radar, station sensors, or call history) can result in more powerful predictions than using single sources in isolation.Finally, we present a methodology for characterizing the different stages of a rainfall event, as performance is expected to vary throughout its evolution. We distinguish five stages based on observed rain in the previous and following hours: The first hour with rain, intermediate hours, the last hour with rain, the hours immediately after the event, and hours without rain. Evaluating all approaches following this framework adds a valuable dimension to the performance analysis and further improves explainability. The results demonstrate that our models outperform the baselines across all event stages, from the initial onset of rain to the hours after precipitation has stopped. This highlights the strong potential of even relatively simple ML pipelines to deliver timely, localized anticipation of weather-related impacts.

  • Open Access

    Are traditional hazard-based weather warnings anticipating local impact?

    by Jordi Morales, Xavier Llort, Andreas Kaltenbrunner, Agata Lapedriza

    Hazard-based weather warnings alert decision makers, responders, and the population about meteorological risks, yet often overlook data on vulnerability and exposure, relying instead on fixed or climatology-based thresholds. This raises questions about their effectiveness at higher resolutions, as weather impacts are ultimately determined by local characteristics. This work addresses this gap by providing the first quantitative and high-resolution assessment of how traditional heavy rain and wind gust warnings correlate with actual impacts, using emergency calls as an impact proxy. The study works at the municipal and hourly resolution across Catalonia, in northeastern Spain, over six years (from October 2018 to February 2025), analysing warnings from two meteorological agencies: the national (AEMET, part of the EU MeteoAlert programme) and regional (SMC). We first compare the spatial distribution of issued warnings and emergency calls, assessing how they align with the rain and wind climatology of the region. Following this, overall warning performance metrics are calculated assuming a real-time setting. Further experiments analyse other critical operational variables, such as lead-time and monthly patterns. Our results reveal that while traditional warnings can identify general impact zones, they are severely limited when considered at local scales due to a high number of false alarms, which can reduce their public use and trust. Based on these findings, we not only provide a solid evidence base on how current weather warnings relate to actual impacts, but also aim to inform and accelerate the transition from traditional, hazard-centric systems towards more actionable and precise impact-based warning systems.

  • Open Access

    Language-Agnostic Modeling of Source Reliability on Wikipedia

    by Jacopo D’IGnazi, Andreas Kaltenbrunner, Yelena Mejova, Michele Tizzani, Kyriaki Kalimeri, Mariano G. Beiró, Pablo Aragón

    Over the last few years, verifying the credibility of information sources has become a fundamental need to combat disinformation. Here, we present a language-agnostic model designed to assess the reliability of web domains as sources in references across multiple language editions of Wikipedia. Utilizing editing activity data, the model evaluates domain reliability within different articles of varying controversiality, such as Climate Change, COVID-19, History, Media, and Biology topics. Crafting features that express domain usage across articles, the model effectively predicts domain reliability, achieving an F1 Macro score of approximately 0.80 for English and other high-resource languages. For mid-resource languages, we achieve 0.65, while the performance of low-resource languages varies. In all cases, the time the domain remains present in the articles (which we dub as permanence ) is one of the most predictive features. We highlight the challenge of maintaining consistent model performance across languages of varying resource levels and demonstrate that adapting models from higher-resource languages can improve performance. We believe these findings can assist Wikipedia editors in their ongoing efforts to verify citations and may offer useful insights for other user-generated content communities.


  • Open Access

    The Personality Trap: How LLMs Embed Bias When Generating Human-Like Personas

    by Jacopo Amidei, Gregorio Ferreira, Mario Muñoz Serrano, Rubén Nieto, Andreas Kaltenbrunner

    This paper examines biases in large language models (LLMs) when generating synthetic populations from responses to personality questionnaires. Using five LLMs, we first assess the representativeness and potential biases in the sociodemographic attributes of the generated personas, as well as their alignment with the intended personality traits. While LLMs successfully reproduce known correlations between personality and sociodemographic variables, all models exhibit pronounced WEIRD (western, educated, industrialized, rich and democratic) biases, favoring young, educated, white, heterosexual, Western individuals with centrist or progressive political views and secular or Christian beliefs. In a second analysis, we manipulate input traits to maximize Neuroticism and Psychoticism scores. Notably, when Psychoticism is maximized, several models produce an overrepresentation of non-binary and LGBTQ+ identities, raising concerns about stereotyping and the potential pathologization of marginalized groups. Our findings highlight both the potential and the risks of using LLMs to generate psychologically grounded synthetic populations.

  • Open Access

    A simple and explainable model for park-and-ride car park occupancy prediction

    by Andreas Kaltenbrunner, Josep Ferrer, David Moreno, Vicenç Gómez

    In a scenario of growing usage of park-and-ride facilities, understanding and predicting car park occupancy is becoming increasingly important. This study presents a model that effectively captures the occupancy patterns of park-and-ride car parks for commuters using truncated normal distributions for vehicle arrival and departure times. The objective is to develop a predictive model with minimal parameters corresponding to commuter behaviour, enabling the estimation of parking saturation and unfulfilled demand. The proposed model successfully identifies the regular, periodic nature of commuter parking behaviour, where vehicles arrive in the morning and depart in the afternoon. It operates using aggregate data, eliminating the need for individual tracking of arrivals and departures. The model’s predictive and nowcasting capabilities are demonstrated through real-world data from car parks in the Barcelona Metropolitan Area. A simple model extension furthermore enables the prediction of when a car park will reach its occupancy limit and estimates the additional spaces required to accommodate such excess demand. Thus, beyond forecasting, the model serves as a valuable tool for evaluating interventions, such as expanding parking capacity, to optimise park-and-ride facilities.


  • Open Access

    Matching GPT-simulated Populations with Real Ones in Psychological Studies—The Case of the EPQR-A Personality Test

    by Gregorio Ferreira, Jacopo Amidei, Rubén Nieto, Andreas Kaltenbrunner

    This article analyzes how well OpenAI’s LLM GPT-4 can emulate different personalities and simulate populations to answer psychological questionnaires similarly to real population samples. For this purpose, we performed different experiments with the Eysenck Personality Questionnaire-Revised Abbreviated (EPQR-A) in three different languages (Spanish, English, and Slovak). The EPQR-A measures personality on four scales: extraversion (E: sociability), neuroticism (N: emotional stability), psychoticism (P: tendency to break social rules, and not having empathy), and lying (L: social desirability). We perform a comparative analysis of the answers of synthetic populations with those of two real population samples of Spanish students as well as the unconditioned baseline personality of GPT. Furthermore, the impact of time (what year the questionnaire is answered), questionnaire language, and student age and gender are analyzed. To our knowledge, this is the first time the EPQR-A test has been used to assess the GPT´s personality and the impact of different language versions and time are measured. Our analysis reveals that GPT-4 exhibits an extroverted, emotionally stable personality with low psychoticism levels and high social desirability. GPT-4 replicates some differences observed in real populations in terms of gender but only partially replicates the results for real populations.