Gerasimos Spanakis
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Gerasimos Spanakis has a total of 159 co-authors. Here are the top co-authors by number of shared publications:
- Anne Roefs
Faculty of Psychology and Neuroscience, Maastricht University, Maastricht, The Netherlands
15 shared publications - Gijs van Dijck
Law & Tech Lab, Maastricht University, Maastricht, Netherlands
13 shared publications Yusuf Can Semerci
Department of Advanced Computing Sciences Maastricht University
13 shared publications- Abderrahmane Issam
Department of Advanced Computing Sciences Maastricht University
8 shared publications - Jan Scholtes
Department of Advanced Computing Sciences Maastricht University
7 shared publications - Mandani Ntekouli
Maastricht University, Maastricht, Netherlands
7 shared publications
Latest Publications
- Open Access
- 16 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-12
Evaluating the User Experience in a Digitally Enhanced AR Theatre Performance with VFX and Auto-Captioning
- 16 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-12
Open AccessEvaluating the User Experience in a Digitally Enhanced AR Theatre Performance with VFX and Auto-Captioning
- 26 October 2026
- Published by Taylor & Francis
- p. 245-265
Influencer Discovery Platforms
Influencer marketing has become a crucial feature of digital marketing strategies. Although academic and industry literature has dedicated a lot of attention to social media platforms, brands, and influencers in this space, there is also a growing economy around new actors and products facilitating a lot of these relationships. This economy reflects data intermediaries known as ‘influencer marketing platforms’ (IMPs). IMPs rely on data from social media platforms’ Application Programming Interfaces (APIs) to create novel data products relevant for industry. Such products can include original indicators for brand safety, classifiers for geographical locations, demographic analytics, or engagement metrics. Social media platforms are already highly opaque infrastructures and given their commercial and more obscure nature in the influencer marketing supply chain, IMPs pose additional transparency issues. There is little to no information about how these products are built, what technologies they use, and what accuracy or biasing issues could give rise to legal and ethical problems. This chapter fills an existing research gap by critically exploring IMPs from a commercial and technological perspective. We first focus on describing this industry as a whole, what role they play in the influencer market, how they can help brands and influencers alike, and how they interact with social media platforms. We then select several large IMPs and give specific illustrations of products relevant for this field (Hypeauditor and Heepsy), while reporting on state-of-the-art technology relevant for these products, based on a literature review in computational studies. Lastly, we critically reflect on whether what is promised is actually matched by reported scientific innovation, and consider what happens when that is not the case.
- 26 October 2026
- Published by Taylor & Francis
- p. 245-265
Influencer Discovery Platforms
Influencer marketing has become a crucial feature of digital marketing strategies. Although academic and industry literature has dedicated a lot of attention to social media platforms, brands, and influencers in this space, there is also a growing economy around new actors and products facilitating a lot of these relationships. This economy reflects data intermediaries known as ‘influencer marketing platforms’ (IMPs). IMPs rely on data from social media platforms’ Application Programming Interfaces (APIs) to create novel data products relevant for industry. Such products can include original indicators for brand safety, classifiers for geographical locations, demographic analytics, or engagement metrics. Social media platforms are already highly opaque infrastructures and given their commercial and more obscure nature in the influencer marketing supply chain, IMPs pose additional transparency issues. There is little to no information about how these products are built, what technologies they use, and what accuracy or biasing issues could give rise to legal and ethical problems. This chapter fills an existing research gap by critically exploring IMPs from a commercial and technological perspective. We first focus on describing this industry as a whole, what role they play in the influencer market, how they can help brands and influencers alike, and how they interact with social media platforms. We then select several large IMPs and give specific illustrations of products relevant for this field (Hypeauditor and Heepsy), while reporting on state-of-the-art technology relevant for these products, based on a literature review in computational studies. Lastly, we critically reflect on whether what is promised is actually matched by reported scientific innovation, and consider what happens when that is not the case.
- Open Access
- 25 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 8184-8201
Is your AI Model Accurate Enough? The Difficult Choices Behind Rigorous AI Development and the EU AI Act
- 25 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 8184-8201
Open AccessIs your AI Model Accurate Enough? The Difficult Choices Behind Rigorous AI Development and the EU AI Act
- Open Access
- 27 April 2026
- Published by Elsevier in SSRN Electronic Journal
Is your AI Model Accurate Enough? The Difficult Choices Behind Rigorous AI Development and the EU AI Act
Technical and legal debates frequently suggest that "accuracy" is an objective, measurable, and purely technical property. We challenge this view, showing that evaluating AI performance fundamentally depends on context-dependent normative decisions. These technonormative choices are crucial for rigorous AI deployment, as they determine which errors are prioritised, how risks are distributed, and how trade-offs between competing objectives are resolved. This paper provides a legal-technical analysis of the choices that shape how accuracy is defined, measured, and assessed, using the 2024 European Union AI Act-which mandates an "appropriate level of accuracy" for high-risk systems-as a primary case study. We identify and analyse four choices central to any robust performance evaluation: (1) selecting metrics, (2) balancing multiple metrics, (3) measuring metrics against representative data, and (4) determining acceptance thresholds. For each choice, we study its relationship to the AI Act's accuracy requirement and associated documentation obligations, show how its technical implementation embeds implicit or explicit assumptions about acceptable risks, errors, and trade-offs, and discuss the implications for the practical implementation of the AI Act by examples and related technical standards. By making the techno-normative dimensions of accuracy explicit, this paper contributes to broader interdisciplinary debates on AI governance and regulation, and offers specific guidance for regulators, auditors, and developers tasked with translating (legal) safety requirements into technical practice.
- 27 April 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessIs your AI Model Accurate Enough? The Difficult Choices Behind Rigorous AI Development and the EU AI Act
Technical and legal debates frequently suggest that "accuracy" is an objective, measurable, and purely technical property. We challenge this view, showing that evaluating AI performance fundamentally depends on context-dependent normative decisions. These technonormative choices are crucial for rigorous AI deployment, as they determine which errors are prioritised, how risks are distributed, and how trade-offs between competing objectives are resolved. This paper provides a legal-technical analysis of the choices that shape how accuracy is defined, measured, and assessed, using the 2024 European Union AI Act-which mandates an "appropriate level of accuracy" for high-risk systems-as a primary case study. We identify and analyse four choices central to any robust performance evaluation: (1) selecting metrics, (2) balancing multiple metrics, (3) measuring metrics against representative data, and (4) determining acceptance thresholds. For each choice, we study its relationship to the AI Act's accuracy requirement and associated documentation obligations, show how its technical implementation embeds implicit or explicit assumptions about acceptable risks, errors, and trade-offs, and discuss the implications for the practical implementation of the AI Act by examples and related technical standards. By making the techno-normative dimensions of accuracy explicit, this paper contributes to broader interdisciplinary debates on AI governance and regulation, and offers specific guidance for regulators, auditors, and developers tasked with translating (legal) safety requirements into technical practice.
- 1 April 2026
- Published by Springer Nature
- p. 86-106
Are Companies Taking AI Risks Seriously? A Systematic Analysis of Companies’ AI Risk Disclosures in SEC 10-K Forms
- 1 April 2026
- Published by Springer Nature
- p. 86-106
Are Companies Taking AI Risks Seriously? A Systematic Analysis of Companies’ AI Risk Disclosures in SEC 10-K Forms
- Open Access
Cross-Modal Robustness Transfer (CMRT): Training Robust Speech Translation Models Using Adversarial Text
End-to-End Speech Translation (E2E-ST) has seen significant advancements, yet current models are primarily benchmarked on curated, "clean" datasets. This overlooks critical real-world challenges, such as morphological robustness to inflectional variations common in non-native or dialectal speech. In this work, we adapt a text-based adversarial attack targeting inflectional morphology to the speech domain and demonstrate that state-of-the-art E2E-ST models are highly vulnerable it. While adversarial training effectively mitigates such risks in text-based tasks, generating high-quality adversarial speech data remains computationally expensive and technically challenging. To address this, we propose Cross-Modal Robustness Transfer (CMRT), a framework that transfers adversarial robustness from the text modality to the speech modality. Our method eliminates the requirement for adversarial speech data during training. Extensive experiments across four language pairs demonstrate that CMRT improves adversarial robustness by an average of more than 3 BLEU points, establishing a new baseline for robust E2E-ST without the overhead of generating adversarial speech.
Open AccessCross-Modal Robustness Transfer (CMRT): Training Robust Speech Translation Models Using Adversarial Text
End-to-End Speech Translation (E2E-ST) has seen significant advancements, yet current models are primarily benchmarked on curated, "clean" datasets. This overlooks critical real-world challenges, such as morphological robustness to inflectional variations common in non-native or dialectal speech. In this work, we adapt a text-based adversarial attack targeting inflectional morphology to the speech domain and demonstrate that state-of-the-art E2E-ST models are highly vulnerable it. While adversarial training effectively mitigates such risks in text-based tasks, generating high-quality adversarial speech data remains computationally expensive and technically challenging. To address this, we propose Cross-Modal Robustness Transfer (CMRT), a framework that transfers adversarial robustness from the text modality to the speech modality. Our method eliminates the requirement for adversarial speech data during training. Extensive experiments across four language pairs demonstrate that CMRT improves adversarial robustness by an average of more than 3 BLEU points, establishing a new baseline for robust E2E-ST without the overhead of generating adversarial speech.
- Open Access
Language Models as Artificial Learners: Investigating Crosslinguistic Influence
Despite the centrality of crosslinguistic influence (CLI) to bilingualism research, human studies often yield conflicting results due to inherent experimental variance. We address these inconsistencies by using language models (LMs) as controlled statistical learners to systematically simulate CLI and isolate its underlying drivers. Specifically, we study the effect of varying the L1 language dominance and the L2 language proficiency, which we manipulate by controlling the L2 age of exposure -- defined as the training step at which the L2 is introduced. Furthermore, we investigate the impact of pretraining on L1 languages with varying syntactic distance from the L2. Using cross-linguistic priming, we analyze how activating L1 structures impacts L2 processing. Our results align with evidence from psycholinguistic studies, confirming that language dominance and proficiency are strong predictors of CLI. We further find that while priming of grammatical structures is bidirectional, the priming of ungrammatical structures is sensitive to language dominance. Finally, we provide mechanistic evidence of CLI in LMs, demonstrating that the L1 is co-activated during L2 processing and directly influences the neural circuitry recruited for the L2. More broadly, our work demonstrates that LMs can serve as a computational framework to inform theories of human CLI.
Open AccessLanguage Models as Artificial Learners: Investigating Crosslinguistic Influence
Despite the centrality of crosslinguistic influence (CLI) to bilingualism research, human studies often yield conflicting results due to inherent experimental variance. We address these inconsistencies by using language models (LMs) as controlled statistical learners to systematically simulate CLI and isolate its underlying drivers. Specifically, we study the effect of varying the L1 language dominance and the L2 language proficiency, which we manipulate by controlling the L2 age of exposure -- defined as the training step at which the L2 is introduced. Furthermore, we investigate the impact of pretraining on L1 languages with varying syntactic distance from the L2. Using cross-linguistic priming, we analyze how activating L1 structures impacts L2 processing. Our results align with evidence from psycholinguistic studies, confirming that language dominance and proficiency are strong predictors of CLI. We further find that while priming of grammatical structures is bidirectional, the priming of ungrammatical structures is sensitive to language dominance. Finally, we provide mechanistic evidence of CLI in LMs, demonstrating that the L1 is co-activated during L2 processing and directly influences the neural circuitry recruited for the L2. More broadly, our work demonstrates that LMs can serve as a computational framework to inform theories of human CLI.
- Open Access
- 12 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 1674-1678
Evaluating the Quality of Counterfactual Explanations in Multivariate Time-Series
- 12 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 1674-1678
Open AccessEvaluating the Quality of Counterfactual Explanations in Multivariate Time-Series
- Open Access
DTW-Align: Bridging the Modality Gap in End-to-End Speech Translation with Dynamic Time Warping Alignment
End-to-End Speech Translation (E2E-ST) is the task of translating source speech directly into target text bypassing the intermediate transcription step. The representation discrepancy between the speech and text modalities has motivated research on what is known as bridging the modality gap. State-of-the-art methods addressed this by aligning speech and text representations on the word or token level. Unfortunately, this requires an alignment tool that is not available for all languages. Although this issue has been addressed by aligning speech and text embeddings using nearest-neighbor similarity search, it does not lead to accurate alignments. In this work, we adapt Dynamic Time Warping (DTW) for aligning speech and text embeddings during training. Our experiments demonstrate the effectiveness of our method in bridging the modality gap in E2E-ST. Compared to previous work, our method produces more accurate alignments and achieves comparable E2E-ST results while being significantly faster. Furthermore, our method outperforms previous work in low resource settings on 5 out of 6 language directions.
Open AccessDTW-Align: Bridging the Modality Gap in End-to-End Speech Translation with Dynamic Time Warping Alignment
End-to-End Speech Translation (E2E-ST) is the task of translating source speech directly into target text bypassing the intermediate transcription step. The representation discrepancy between the speech and text modalities has motivated research on what is known as bridging the modality gap. State-of-the-art methods addressed this by aligning speech and text representations on the word or token level. Unfortunately, this requires an alignment tool that is not available for all languages. Although this issue has been addressed by aligning speech and text embeddings using nearest-neighbor similarity search, it does not lead to accurate alignments. In this work, we adapt Dynamic Time Warping (DTW) for aligning speech and text embeddings during training. Our experiments demonstrate the effectiveness of our method in bridging the modality gap in E2E-ST. Compared to previous work, our method produces more accurate alignments and achieves comparable E2E-ST results while being significantly faster. Furthermore, our method outperforms previous work in low resource settings on 5 out of 6 language directions.
- Open Access
You Are What You Train: Effects of Data Composition on Training Context-aware Machine Translation Models
Achieving human-level translations requires leveraging context to ensure coherence and handle complex phenomena like pronoun disambiguation. Sparsity of contextually rich examples in the standard training data has been hypothesized as the reason for the difficulty of context utilization. In this work, we systematically validate this claim in both single- and multilingual settings by constructing training datasets with a controlled proportions of contextually relevant examples. We demonstrate a strong association between training data sparsity and model performance confirming sparsity as a key bottleneck. Importantly, we reveal that improvements in one contextual phenomenon do no generalize to others. While we observe some cross-lingual transfer, it is not significantly higher between languages within the same sub-family. Finally, we propose and empirically evaluate two training strategies designed to leverage the available data. These strategies improve context utilization, resulting in accuracy gains of up to 6 and 8 percentage points on the ctxPro evaluation in single- and multilingual settings respectively.
Open AccessYou Are What You Train: Effects of Data Composition on Training Context-aware Machine Translation Models
Achieving human-level translations requires leveraging context to ensure coherence and handle complex phenomena like pronoun disambiguation. Sparsity of contextually rich examples in the standard training data has been hypothesized as the reason for the difficulty of context utilization. In this work, we systematically validate this claim in both single- and multilingual settings by constructing training datasets with a controlled proportions of contextually relevant examples. We demonstrate a strong association between training data sparsity and model performance confirming sparsity as a key bottleneck. Importantly, we reveal that improvements in one contextual phenomenon do no generalize to others. While we observe some cross-lingual transfer, it is not significantly higher between languages within the same sub-family. Finally, we propose and empirically evaluate two training strategies designed to leverage the available data. These strategies improve context utilization, resulting in accuracy gains of up to 6 and 8 percentage points on the ctxPro evaluation in single- and multilingual settings respectively.