Tobias Vente
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Tobias Vente has a total of 20 co-authors. Here are the top co-authors by number of shared publications:
- Joeran Beel
Department of Electrical Engineering and Computer Science, University of Siegen, Siegen, Germany
11 shared publications - Lukas Wegmeth
Department of Electrical Engineering and Computer Science, University of Siegen, Siegen, Germany
10 shared publications - Alan Said
University of Gothenburg
4 shared publications - Lennart Purucker
University of Freiburg
2 shared publications - Abdullah Abbas
Intelligent Systems Group, University of Siegen, Siegen, Germany
1 shared publication - Alan Said
University of Gothenburg, Sweden
1 shared publication
Latest Publications
- Open Access
- 27 September 2026
- Published by Association for Computing Machinery (ACM)
- p. 1802-1804
FINALLY: A Dataset Recommender for RecSys Experiments
- 27 September 2026
- Published by Association for Computing Machinery (ACM)
- p. 1802-1804
Open AccessFINALLY: A Dataset Recommender for RecSys Experiments
- 8 July 2026
- Published by University of Antwerp
Automated recommender systems : advancing pipeline design, model selection, and environmental sustainability
The unprecedented growth of digital data underscores the critical need for recommender systems (RecSys) to filter and organize relevant content. However, despite the success of RecSys, designing and developing effective research pipelines for RecSys remains a labor- and resource-intensive task. At the foundation of RecSys research experiments lies the critical task of identifying appropriate datasets for empirical evaluation. To simplify this, researchers frequently default to a narrow subset of popular datasets, which undermines robustness by constraining evaluations and risking overfitting. To ensure more robust evaluations, this dissertation first tackles the dataset selection problem. By proposing novel, data-driven methodologies and visualization tools, we enable practitioners to easily compare and select RecSys datasets. This mitigates the risk of cherry-picking and ensures generalizable benchmarking. To overcome the manual engineering of the RecSys research pipeline, this work investigates the application of automated machine learning (AutoML) principles to the recommendation domain; formally known as AutoRecSys. Motivated by AutoML’s success in streamlining model development, we specifically focus on solving the automated model selection problem. We evaluate the potential of out-of-the-box AutoML tools and propose the first fully automated AutoRecSys framework capable of simultaneously executing automated algorithm and hyperparameter selection. Additionally, we provide the first systematic analysis of ensembling RecSys models to improve predictive performance and system robustness. Beyond automation and empirical evaluation, we conduct the first comprehensive investigation of the environmental impact of RecSys experiments. Our analysis reveals a concerning ecological trajectory: modern deep learning-based research papers emit approximately 42 times more CO2 equivalents than traditional approaches used 10 years ago. Given this trend, applying computationally demanding automated frameworks might seem counterintuitive. To summarize, this dissertation lays a comprehensive foundation for future research. This work advances data-driven dataset selection, pioneers AutoRecSys model selection frameworks, and quantifies ecological footprints. In doing so, it empowers the RecSys community to build highly performant, automated systems while actively minimizing their environmental impact for a sustainable future.
- 8 July 2026
- Published by University of Antwerp
Automated recommender systems : advancing pipeline design, model selection, and environmental sustainability
The unprecedented growth of digital data underscores the critical need for recommender systems (RecSys) to filter and organize relevant content. However, despite the success of RecSys, designing and developing effective research pipelines for RecSys remains a labor- and resource-intensive task. At the foundation of RecSys research experiments lies the critical task of identifying appropriate datasets for empirical evaluation. To simplify this, researchers frequently default to a narrow subset of popular datasets, which undermines robustness by constraining evaluations and risking overfitting. To ensure more robust evaluations, this dissertation first tackles the dataset selection problem. By proposing novel, data-driven methodologies and visualization tools, we enable practitioners to easily compare and select RecSys datasets. This mitigates the risk of cherry-picking and ensures generalizable benchmarking. To overcome the manual engineering of the RecSys research pipeline, this work investigates the application of automated machine learning (AutoML) principles to the recommendation domain; formally known as AutoRecSys. Motivated by AutoML’s success in streamlining model development, we specifically focus on solving the automated model selection problem. We evaluate the potential of out-of-the-box AutoML tools and propose the first fully automated AutoRecSys framework capable of simultaneously executing automated algorithm and hyperparameter selection. Additionally, we provide the first systematic analysis of ensembling RecSys models to improve predictive performance and system robustness. Beyond automation and empirical evaluation, we conduct the first comprehensive investigation of the environmental impact of RecSys experiments. Our analysis reveals a concerning ecological trajectory: modern deep learning-based research papers emit approximately 42 times more CO2 equivalents than traditional approaches used 10 years ago. Given this trend, applying computationally demanding automated frameworks might seem counterintuitive. To summarize, this dissertation lays a comprehensive foundation for future research. This work advances data-driven dataset selection, pioneers AutoRecSys model selection frameworks, and quantifies ecological footprints. In doing so, it empowers the RecSys community to build highly performant, automated systems while actively minimizing their environmental impact for a sustainable future.
- Open Access
- 25 September 2025
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Recommender Systems
Green Recommender Systems: Understanding and Minimizing the Carbon Footprint of AI-Powered Personalization
As global warming soars, the need to assess and reduce the environmental impact of recommender systems is becoming increasingly urgent. Despite this, the recommender systems community hardly understands, addresses, and evaluates the environmental impact of their work. In this study, we examine the environmental impact of recommender systems research by reproducing typical experimental pipelines. Based on our results, we provide guidelines for researchers and practitioners on how to minimize the environmental footprint of their work and implement green recommender systems — recommender systems designed to minimize their energy consumption and carbon footprint. Our analysis covers 79 papers from the 2013 and 2023 ACM RecSys conferences, comparing traditional “good old-fashioned AI” models with modern deep learning models. We designed and reproduced representative experimental pipelines for both years, measuring energy consumption using a hardware energy meter and converting it into CO 2 equivalents. Our results show that papers utilizing deep learning models emit approximately 42 times more CO 2 equivalents than papers using traditional models. On average, a single deep learning-based paper generates 2,909 kilograms of CO 2 equivalents — more than the carbon emissions of a person flying from New York City to Melbourne or the amount of CO 2 sequestered by one tree over 260 years. This work underscores the urgent need for the recommender systems and wider machine learning communities to adopt green AI principles, balancing algorithmic advancements and environmental responsibility to build a sustainable future with AI-powered personalization.
- 25 September 2025
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Recommender Systems
Open AccessGreen Recommender Systems: Understanding and Minimizing the Carbon Footprint of AI-Powered Personalization
As global warming soars, the need to assess and reduce the environmental impact of recommender systems is becoming increasingly urgent. Despite this, the recommender systems community hardly understands, addresses, and evaluates the environmental impact of their work. In this study, we examine the environmental impact of recommender systems research by reproducing typical experimental pipelines. Based on our results, we provide guidelines for researchers and practitioners on how to minimize the environmental footprint of their work and implement green recommender systems — recommender systems designed to minimize their energy consumption and carbon footprint. Our analysis covers 79 papers from the 2013 and 2023 ACM RecSys conferences, comparing traditional “good old-fashioned AI” models with modern deep learning models. We designed and reproduced representative experimental pipelines for both years, measuring energy consumption using a hardware energy meter and converting it into CO 2 equivalents. Our results show that papers utilizing deep learning models emit approximately 42 times more CO 2 equivalents than papers using traditional models. On average, a single deep learning-based paper generates 2,909 kilograms of CO 2 equivalents — more than the carbon emissions of a person flying from New York City to Melbourne or the amount of CO 2 sequestered by one tree over 260 years. This work underscores the urgent need for the recommender systems and wider machine learning communities to adopt green AI principles, balancing algorithmic advancements and environmental responsibility to build a sustainable future with AI-powered personalization.
- Open Access
- 22 September 2025
- Published by Association for Computing Machinery (ACM)
- p. 1322-1324
APS Explorer: Navigating Algorithm Performance Spaces for Informed Dataset Selection
- 22 September 2025
- Published by Association for Computing Machinery (ACM)
- p. 1322-1324
Open AccessAPS Explorer: Navigating Algorithm Performance Spaces for Informed Dataset Selection
- Open Access
Green Recommender Systems: Understanding and Minimizing the Carbon Footprint of AI-Powered Personalization
As global warming soars, the need to assess and reduce the environmental impact of recommender systems is becoming increasingly urgent. Despite this, the recommender systems community hardly understands, addresses, and evaluates the environmental impact of their work. In this study, we examine the environmental impact of recommender systems research by reproducing typical experimental pipelines. Based on our results, we provide guidelines for researchers and practitioners on how to minimize the environmental footprint of their work and implement green recommender systems - recommender systems designed to minimize their energy consumption and carbon footprint. Our analysis covers 79 papers from the 2013 and 2023 ACM RecSys conferences, comparing traditional "good old-fashioned AI" models with modern deep learning models. We designed and reproduced representative experimental pipelines for both years, measuring energy consumption using a hardware energy meter and converting it into CO2 equivalents. Our results show that papers utilizing deep learning models emit approximately 42 times more CO2 equivalents than papers using traditional models. On average, a single deep learning-based paper generates 2,909 kilograms of CO2 equivalents - more than the carbon emissions of a person flying from New York City to Melbourne or the amount of CO2 sequestered by one tree over 260 years. This work underscores the urgent need for the recommender systems and wider machine learning communities to adopt green AI principles, balancing algorithmic advancements and environmental responsibility to build a sustainable future with AI-powered personalization.
Open AccessGreen Recommender Systems: Understanding and Minimizing the Carbon Footprint of AI-Powered Personalization
As global warming soars, the need to assess and reduce the environmental impact of recommender systems is becoming increasingly urgent. Despite this, the recommender systems community hardly understands, addresses, and evaluates the environmental impact of their work. In this study, we examine the environmental impact of recommender systems research by reproducing typical experimental pipelines. Based on our results, we provide guidelines for researchers and practitioners on how to minimize the environmental footprint of their work and implement green recommender systems - recommender systems designed to minimize their energy consumption and carbon footprint. Our analysis covers 79 papers from the 2013 and 2023 ACM RecSys conferences, comparing traditional "good old-fashioned AI" models with modern deep learning models. We designed and reproduced representative experimental pipelines for both years, measuring energy consumption using a hardware energy meter and converting it into CO2 equivalents. Our results show that papers utilizing deep learning models emit approximately 42 times more CO2 equivalents than papers using traditional models. On average, a single deep learning-based paper generates 2,909 kilograms of CO2 equivalents - more than the carbon emissions of a person flying from New York City to Melbourne or the amount of CO2 sequestered by one tree over 260 years. This work underscores the urgent need for the recommender systems and wider machine learning communities to adopt green AI principles, balancing algorithmic advancements and environmental responsibility to build a sustainable future with AI-powered personalization.
- Open Access
- 31 August 2025
- Published by Association for Computing Machinery (ACM)
- p. 4-15
Automated Learning of Software Configuration Spaces is not Easy
- 31 August 2025
- Published by Association for Computing Machinery (ACM)
- p. 4-15
Open AccessAutomated Learning of Software Configuration Spaces is not Easy
- Open Access
- 12 June 2025
- Published by Association for Computing Machinery (ACM)
- p. 371-378
The Potential of AutoML for Recommender Systems
- 12 June 2025
- Published by Association for Computing Machinery (ACM)
- p. 371-378
Open AccessThe Potential of AutoML for Recommender Systems
- 9 April 2025
- Published by Springer Nature
- p. 90-97
e-Fold Cross-Validation for Recommender-System Evaluation
- 9 April 2025
- Published by Springer Nature
- p. 90-97
e-Fold Cross-Validation for Recommender-System Evaluation
- 3 April 2025
- Published by Springer Nature
- p. 30-34
Checky, the Paper-Submission Checklist Generator for Authors, Reviewers and LLMs
- 3 April 2025
- Published by Springer Nature
- p. 30-34
Checky, the Paper-Submission Checklist Generator for Authors, Reviewers and LLMs