Antal van Den Bosch
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Antal van Den Bosch has a total of 152 co-authors. Here are the top co-authors by number of shared publications:
- Walter Daelemans
Department of Linguistics, University of Antwerp, Antwerp, Belgium
13 shared publications - Iris Hendrickx
Centre for Language Studies Centre for Language and Speech Technology Radboud University Nijmegen The Netherlands
8 shared publications Toine Bogers
Tilburg University, The Netherlands
8 shared publicationsWalter Daelemans
University of Antwerp, Antwerp, Belgium
6 shared publicationsSander Wubben
Tilburg University, Tilburg, Netherlands
5 shared publicationsTon Weijters
Department of I&T, Faculty of Technology Management, Eindhoven University of Technology, P.O. Box 513, 5600 MB Eindhoven, The Netherlands
5 shared publications
Latest Publications
- Open Access
- 29 June 2026
- Published by Taylor & Francis in Discourse Processes
- Vol. 63 (6-7) , 1-25
The implicit guides to convincingness intuitions: how emotional valence, emotional arousal, and processing fluency relate to perceived and actual convincingness
- 29 June 2026
- Published by Taylor & Francis in Discourse Processes
- Vol. 63 (6-7) , 1-25
Open AccessThe implicit guides to convincingness intuitions: how emotional valence, emotional arousal, and processing fluency relate to perceived and actual convincingness
- Open Access
- 29 December 2025
- Published by SAGE Publications in Social Science Computer Review
Who Are the Online Commenters? A Large-Scale Representative Survey to Explore the Identity and Motivation of Online News Commenters in Comparison to Non-Commenters
To better understand the demographic composition of people participating in commenting sections beneath online news articles, we conducted a large-scale survey ( n = 5,490) with a panel that is representative of the Dutch population – the LISS panel. We combined these data with demographic background variables and previously collected data on political views and values, to provide a detailed description of the identity of online news commenters in comparison to non-commenters. Our results show that the group of commenters contain more men (55%), and the age group of 45–54 years old has the largest share of commenters (18% for men, 13% for women). Furthermore, we found little to no differences for education levels, income, location, political preferences, and cultural background, suggesting that there is no striking overrepresentation of specific groups among online commenters in general. However, when looking at the profiles of online commenters as a function of the topic and platform of discussion, differences start to emerge for gender, age, and education levels. We found no differences related to age and gender distributions for those with a higher commenting frequency, but a higher frequency does go hand in hand with more support for national populist and far-right political parties and a lower confidence in political parties.
- 29 December 2025
- Published by SAGE Publications in Social Science Computer Review
Open AccessWho Are the Online Commenters? A Large-Scale Representative Survey to Explore the Identity and Motivation of Online News Commenters in Comparison to Non-Commenters
To better understand the demographic composition of people participating in commenting sections beneath online news articles, we conducted a large-scale survey ( n = 5,490) with a panel that is representative of the Dutch population – the LISS panel. We combined these data with demographic background variables and previously collected data on political views and values, to provide a detailed description of the identity of online news commenters in comparison to non-commenters. Our results show that the group of commenters contain more men (55%), and the age group of 45–54 years old has the largest share of commenters (18% for men, 13% for women). Furthermore, we found little to no differences for education levels, income, location, political preferences, and cultural background, suggesting that there is no striking overrepresentation of specific groups among online commenters in general. However, when looking at the profiles of online commenters as a function of the topic and platform of discussion, differences start to emerge for gender, age, and education levels. We found no differences related to age and gender distributions for those with a higher commenting frequency, but a higher frequency does go hand in hand with more support for national populist and far-right political parties and a lower confidence in political parties.
- Open Access
Memory-based Language Models: An Efficient, Explainable, and Eco-friendly Approach to Large Language Modeling
We present memory-based language modeling as an efficient, eco-friendly alternative to deep neural network-based language modeling. It offers log-linearly scalable next-token prediction performance and strong memorization capabilities. Implementing fast approximations of k-nearest neighbor classification, memory-based language modeling leaves a relatively small ecological footprint both in training and in inference mode, as it relies fully on CPUs and attains low token latencies. Its internal workings are simple and fully transparent. We compare our implementation of memory-based language modeling, OLIFANT, with GPT-2 and GPT-Neo on next-token prediction accuracy, estimated emissions and speeds, and offer some deeper analyses of the model.
Open AccessMemory-based Language Models: An Efficient, Explainable, and Eco-friendly Approach to Large Language Modeling
We present memory-based language modeling as an efficient, eco-friendly alternative to deep neural network-based language modeling. It offers log-linearly scalable next-token prediction performance and strong memorization capabilities. Implementing fast approximations of k-nearest neighbor classification, memory-based language modeling leaves a relatively small ecological footprint both in training and in inference mode, as it relies fully on CPUs and attains low token latencies. Its internal workings are simple and fully transparent. We compare our implementation of memory-based language modeling, OLIFANT, with GPT-2 and GPT-Neo on next-token prediction accuracy, estimated emissions and speeds, and offer some deeper analyses of the model.
- Open Access
Improving the Inclusivity of Dutch Speech Recognition by Fine-tuning Whisper on the JASMIN-CGN Corpus
We test and study the variation in speech recognition of fine-tuned versions of the Whisper model on child, elderly and non-native Dutch speech from the JASMIN-CGN corpus. Our primary goal is to evaluate how speakers' age and linguistic background influence Whisper's performance. Whisper achieves varying Word Error Rates (WER) when fine-tuned on subpopulations of specific ages and linguistic backgrounds. Fine-tuned performance is remarkably better than zero-shot performance, achieving a relative reduction in WER of 81% for native children, 72% for non-native children, 67% for non-native adults, and 65% for native elderly people. Our findings underscore the importance of training speech recognition models like Whisper on underrepresented subpopulations such as children, the elderly, and non-native speakers.
Open AccessImproving the Inclusivity of Dutch Speech Recognition by Fine-tuning Whisper on the JASMIN-CGN Corpus
We test and study the variation in speech recognition of fine-tuned versions of the Whisper model on child, elderly and non-native Dutch speech from the JASMIN-CGN corpus. Our primary goal is to evaluate how speakers' age and linguistic background influence Whisper's performance. Whisper achieves varying Word Error Rates (WER) when fine-tuned on subpopulations of specific ages and linguistic backgrounds. Fine-tuned performance is remarkably better than zero-shot performance, achieving a relative reduction in WER of 81% for native children, 72% for non-native children, 67% for non-native adults, and 65% for native elderly people. Our findings underscore the importance of training speech recognition models like Whisper on underrepresented subpopulations such as children, the elderly, and non-native speakers.
- 25 January 2025
- Published by Springer Nature
- p. 65-74
The Impact of Featuring Comments in Online Discussions
- 25 January 2025
- Published by Springer Nature
- p. 65-74
The Impact of Featuring Comments in Online Discussions
- Open Access
The Impact of Featuring Comments in Online Discussions
A widespread moderation strategy by online news platforms is to feature what the platform deems high quality comments, usually called editor picks or featured comments. In this paper, we compare online discussions of news articles in which certain comments are featured, versus discussions in which no comments are featured. We measure the impact of featuring comments on the discussion, by estimating and comparing the quality of discussions from the perspective of the user base and the platform itself. Our analysis shows that the impact on discussion quality is limited. However, we do observe an increase in discussion activity after the first comments are featured by moderators, suggesting that the moderation strategy might be used to increase user engagement and to postpone the natural decline in user activity over time.
Open AccessThe Impact of Featuring Comments in Online Discussions
A widespread moderation strategy by online news platforms is to feature what the platform deems high quality comments, usually called editor picks or featured comments. In this paper, we compare online discussions of news articles in which certain comments are featured, versus discussions in which no comments are featured. We measure the impact of featuring comments on the discussion, by estimating and comparing the quality of discussions from the perspective of the user base and the platform itself. Our analysis shows that the impact on discussion quality is limited. However, we do observe an increase in discussion activity after the first comments are featured by moderators, suggesting that the moderation strategy might be used to increase user engagement and to postpone the natural decline in user activity over time.
- Open Access
- 26 March 2024
- Published by Dutch Journal of Applied Linguistics (DuJAL) in Dutch Journal of Applied Linguistics
- Vol. 13
Taming our Wild Data
Dutch Journal of Applied Linguistics. Official journal of the Dutch Association of Applied Linguistics (Anéla)
- 26 March 2024
- Published by Dutch Journal of Applied Linguistics (DuJAL) in Dutch Journal of Applied Linguistics
- Vol. 13
Open AccessTaming our Wild Data
Dutch Journal of Applied Linguistics. Official journal of the Dutch Association of Applied Linguistics (Anéla)
- Open Access
- 1 January 2024
- Published by Association for Computational Linguistics (ACL)
- p. 234-243
A Bayesian Quantification of Aporophobia and the Aggravating Effect of Low–Wealth Contexts on Stigmatization
Ryan Brate, Marieke Van Erp, Antal Van Den Bosch. Proceedings of the 8th Workshop on Online Abuse and Harms (WOAH 2024). 2024.
- 1 January 2024
- Published by Association for Computational Linguistics (ACL)
- p. 234-243
Open AccessA Bayesian Quantification of Aporophobia and the Aggravating Effect of Low–Wealth Contexts on Stigmatization
Ryan Brate, Marieke Van Erp, Antal Van Den Bosch. Proceedings of the 8th Workshop on Online Abuse and Harms (WOAH 2024). 2024.
- Open Access
Hybrid moderation in the newsroom: Recommending featured posts to content moderators
Online news outlets are grappling with the moderation of user-generated content within their comment section. We present a recommender system based on ranking class probabilities to support and empower the moderator in choosing featured posts, a time-consuming task. By combining user and textual content features we obtain an optimal classification F1-score of 0.44 on the test set. Furthermore, we observe an optimum mean NDCG@5 of 0.87 on a large set of validation articles. As an expert evaluation, content moderators assessed the output of a random selection of articles by choosing comments to feature based on the recommendations, which resulted in a NDCG score of 0.83. We conclude that first, adding text features yields the best score and second, while choosing featured content remains somewhat subjective, content moderators found suitable comments in all but one evaluated recommendations. We end the paper by analyzing our best-performing model, a step towards transparency and explainability in hybrid content moderation.
Open AccessHybrid moderation in the newsroom: Recommending featured posts to content moderators
Online news outlets are grappling with the moderation of user-generated content within their comment section. We present a recommender system based on ranking class probabilities to support and empower the moderator in choosing featured posts, a time-consuming task. By combining user and textual content features we obtain an optimal classification F1-score of 0.44 on the test set. Furthermore, we observe an optimum mean NDCG@5 of 0.87 on a large set of validation articles. As an expert evaluation, content moderators assessed the output of a random selection of articles by choosing comments to feature based on the recommendations, which resulted in a NDCG score of 0.83. We conclude that first, adding text features yields the best score and second, while choosing featured content remains somewhat subjective, content moderators found suitable comments in all but one evaluated recommendations. We end the paper by analyzing our best-performing model, a step towards transparency and explainability in hybrid content moderation.
- Open Access
- 4 November 2022
- Published by Elsevier in Brain and Language
- Vol. 235, 105197
Speech register influences listeners’ word expectations
- 4 November 2022
- Published by Elsevier in Brain and Language
- Vol. 235, 105197
Open AccessSpeech register influences listeners’ word expectations