@article{56190,
  abstract     = {{This study investigates the potential of using advanced conversational artificial intelligence (AI) to help people understand complex AI systems. In line with conversation-analytic research, we view the participatory role of AI as dynamically unfolding in a situation rather than being predetermined by its architecture. To study user sensemaking of intransparent AI systems, we set up a naturalistic encounter between human participants and two AI systems developed in-house: a reinforcement learning simulation and a GPT-4-based explainer chatbot. Our results reveal that an explainer-AI only truly functions as such when participants actively engage with it as a co-constructive agent. Both the interface’s spatial configuration and the asynchronous temporal nature of the explainer AI – combined with the users’ presuppositions about its role – contribute to the decision whether to treat the AI as a dialogical co-participant in the interaction. Participants establish evidentiality conventions and sensemaking procedures that may diverge from a system’s intended design or function.}},
  author       = {{Klowait, Nils and Erofeeva, Maria and Lenke, Michael and Horwath, Ilona and Buschmeier, Hendrik}},
  journal      = {{Discourse & Communication}},
  number       = {{6}},
  pages        = {{917--930}},
  publisher    = {{Sage}},
  title        = {{{Can AI explain AI? Interactive co-construction of explanations among human and artificial agents}}},
  doi          = {{10.1177/17504813241267069}},
  volume       = {{18}},
  year         = {{2024}},
}

@inproceedings{58224,
  author       = {{Kenneweg, Philip and Kenneweg, Tristan and Fumagalli, Fabian and Hammer, Barbara}},
  booktitle    = {{2024 International Joint Conference on Neural Networks (IJCNN)}},
  keywords     = {{Training, Schedules, Codes, Search methods, Source coding, Computer architecture, Transformers}},
  pages        = {{1--8}},
  title        = {{{No learning rates needed: Introducing SALSA - Stable Armijo Line Search Adaptation}}},
  doi          = {{10.1109/IJCNN60899.2024.10650124}},
  year         = {{2024}},
}

@inproceedings{53073,
  abstract     = {{While shallow decision trees may be interpretable, larger ensemble models like gradient-boosted trees, which often set the state of the art in machine learning problems involving tabular data, still remain black box models. As a remedy, the Shapley value (SV) is a well-known concept in explainable artificial intelligence (XAI) research for quantifying additive feature attributions of predictions. The model-specific TreeSHAP methodology solves the exponential complexity for retrieving exact SVs from tree-based models. Expanding beyond individual feature attribution, Shapley interactions reveal the impact of intricate feature interactions of any order. In this work, we present TreeSHAP-IQ, an efficient method to compute any-order additive Shapley interactions for predictions of tree-based models. TreeSHAP-IQ is supported by a mathematical framework that exploits polynomial arithmetic to compute the interaction scores in a single recursive traversal of the tree, akin to Linear TreeSHAP. We apply TreeSHAP-IQ on state-of-the-art tree ensembles and explore interactions on well-established benchmark datasets.}},
  author       = {{Muschalik, Maximilian and Fumagalli, Fabian and Hammer, Barbara and Huellermeier, Eyke}},
  booktitle    = {{Proceedings of the AAAI Conference on Artificial Intelligence (AAAI)}},
  issn         = {{2374-3468}},
  keywords     = {{Explainable Artificial Intelligence}},
  number       = {{13}},
  pages        = {{14388--14396}},
  title        = {{{Beyond TreeSHAP: Efficient Computation of Any-Order Shapley Interactions for Tree Ensembles}}},
  doi          = {{10.1609/aaai.v38i13.29352}},
  volume       = {{38}},
  year         = {{2024}},
}

@inproceedings{55311,
  abstract     = {{Addressing the limitations of individual attribution scores via the Shapley value (SV), the field of explainable AI (XAI) has recently explored intricate interactions of features or data points. In particular, extensions of the SV, such as the Shapley Interaction Index (SII), have been proposed as a measure to still benefit from the axiomatic basis of the SV. However, similar to the SV, their exact computation remains computationally prohibitive. Hence, we propose with SVARM-IQ a sampling-based approach to efficiently approximate Shapley-based interaction indices of any order. SVARM-IQ can be applied to a broad class of interaction indices, including the SII, by leveraging a novel stratified representation. We provide non-asymptotic theoretical guarantees on its approximation quality and empirically demonstrate that SVARM-IQ achieves state-of-the-art estimation results in practical XAI scenarios on different model classes and application domains.}},
  author       = {{Kolpaczki, Patrick and Muschalik, Maximilian and Fumagalli, Fabian and Hammer, Barbara and Huellermeier, Eyke}},
  booktitle    = {{Proceedings of The 27th International Conference on Artificial Intelligence and Statistics (AISTATS)}},
  pages        = {{3520–3528}},
  publisher    = {{PMLR}},
  title        = {{{SVARM-IQ: Efficient Approximation of Any-order Shapley Interactions through Stratification}}},
  volume       = {{238}},
  year         = {{2024}},
}

@inproceedings{58223,
  abstract     = {{The Shapley value (SV) is a prevalent approach of allocating credit to machine learning (ML) entities to understand black box ML models. Enriching such interpretations with higher-order interactions is inevitable for complex systems, where the Shapley Interaction Index (SII) is a direct axiomatic extension of the SV. While it is well-known that the SV yields an optimal approximation of any game via a weighted least square (WLS) objective, an extension of this result to SII has been a long-standing open problem, which even led to the proposal of an alternative index. In this work, we characterize higher-order SII as a solution to a WLS problem, which constructs an optimal approximation via SII and k-Shapley values (k-SII). We prove this representation for the SV and pairwise SII and give empirically validated conjectures for higher orders. As a result, we propose KernelSHAP-IQ, a direct extension of KernelSHAP for SII, and demonstrate state-of-the-art performance for feature interactions.}},
  author       = {{Fumagalli, Fabian and Muschalik, Maximilian and Kolpaczki, Patrick and Hüllermeier, Eyke and Hammer, Barbara}},
  booktitle    = {{Proceedings of the 41st International Conference on Machine Learning (ICML)}},
  pages        = {{14308–14342}},
  publisher    = {{PMLR}},
  title        = {{{KernelSHAP-IQ: Weighted Least Square Optimization for Shapley Interactions}}},
  volume       = {{235}},
  year         = {{2024}},
}

@inproceedings{61228,
  author       = {{Muschalik, Maximilian and Baniecki, Hubert and Fumagalli, Fabian and Kolpaczki, Patrick and Hammer, Barbara and Huellermeier, Eyke}},
  booktitle    = {{Advances in Neural Information Processing Systems (NeurIPS)}},
  pages        = {{130324–130357}},
  title        = {{{shapiq: Shapley interactions for machine learning}}},
  volume       = {{37}},
  year         = {{2024}},
}

@inproceedings{61230,
  author       = {{Kolpaczki, Patrick and Bengs, Viktor and Muschalik, Maximilian and Hüllermeier, Eyke}},
  booktitle    = {{Proceedings of the AAAI conference on Artificial Intelligence (AAAI)}},
  number       = {{12}},
  pages        = {{13246–13255}},
  title        = {{{Approximating the shapley value without marginal contributions}}},
  volume       = {{38}},
  year         = {{2024}},
}

@inproceedings{61176,
  abstract     = {{We revisit the phenomenon of syntactic complexity convergence in conversational interaction, originally found for English dialogue, which has theoretical implication for dialogical concepts such as mutual understanding. We use a modified metric to quantify syntactic complexity based on dependency parsing. The results show that syntactic complexity convergence can be statistically confirmed in one of three selected German datasets that were analysed. Given that the dataset which shows such convergence is much larger than the other two selected datasets, the empirical results indicate a certain degree of linguistic generality of syntactic complexity convergence in conversational interaction. We also found a different type of syntactic complexity convergence in one of the datasets while further investigation is still necessary.}},
  author       = {{Wang, Yu and Buschmeier, Hendrik}},
  booktitle    = {{Proceedings of the 20th Conference on Natural Language Processing (KONVENS 2024)}},
  location     = {{Vienna, Austria}},
  pages        = {{75–80}},
  title        = {{{Revisiting the phenomenon of syntactic complexity convergence on German dialogue data}}},
  year         = {{2024}},
}

@inproceedings{55911,
  abstract     = {{According to the Entropy Rate Constancy (ERC) principle, the information density of a text is approximately constant over its length. Whether this principle also applies to nonverbal communication signals is still under investigation. We perform empirical analyses of video-recorded dialogue data and investigate whether listener gaze, as an important nonverbal communication signal, adheres to the ERC principle. Results show (1) that the ERC principle holds for listener gaze; and (2) that the two linguistic factors syntactic complexity and turn transition potential are weakly correlated with local entropy of listener gaze.}},
  author       = {{Wang, Yu and Xu, Yang and Skantze, Gabriel and Buschmeier, Hendrik}},
  booktitle    = {{Findings of the Association for Computational Linguistics ACL 2024}},
  location     = {{Bangkok, Thailand}},
  pages        = {{3533–3545}},
  title        = {{{How much does nonverbal communication conform to entropy rate constancy?: A case study on listener gaze in interaction}}},
  year         = {{2024}},
}

@inproceedings{56314,
  author       = {{Riechmann, Alina Naomi and Buschmeier, Hendrik}},
  booktitle    = {{Book of Abstracts of the 2nd International Multimodal Communication Symposium}},
  location     = {{Frankfurt am Main, Germany}},
  pages        = {{38–39}},
  title        = {{{Automatic reconstruction of dialogue participants’ coordinating gaze behavior from multiple camera perspectives}}},
  year         = {{2024}},
}

@inproceedings{58722,
  abstract     = {{Dialects introduce syntactic and lexical variations in language that occur in regional or social groups. Most NLP methods are not sensitive to such variations. This may lead to unfair behavior of the methods, conveying negative bias towards dialect speakers. While previous work has studied dialect-related fairness for aspects like hate speech, other aspects of biased language, such as lewdness, remain fully unexplored. To fill this gap, we investigate performance disparities between dialects in the detection of five aspects of biased language and how to mitigate them. To alleviate bias, we present a multitask learning approach that models dialect language as an auxiliary task to incorporate syntactic and lexical variations. In our experiments with African-American English dialect, we provide empirical evidence that complementing common learning approaches with dialect modeling improves their fairness. Furthermore, the results suggest that multitask learning achieves state-of-the-art performance and helps to detect properties of biased language more reliably.}},
  author       = {{Spliethöver, Maximilian and Menon, Sai Nikhil and Wachsmuth, Henning}},
  booktitle    = {{Findings of the Association for Computational Linguistics: ACL 2024}},
  editor       = {{Ku, Lun-Wei and Martins, Andre and Srikumar, Vivek}},
  pages        = {{9294–9313}},
  publisher    = {{Association for Computational Linguistics}},
  title        = {{{Disentangling Dialect from Social Bias via Multitask Learning to Improve Fairness}}},
  doi          = {{10.18653/v1/2024.findings-acl.553}},
  year         = {{2024}},
}

@inproceedings{61273,
  abstract     = {{In human-machine explanation interactions, such as tutoring systems or customer support chatbots, it is important for the machine explainer to infer the human user's understanding.  Nonverbal signals play an important role for expressing mental states like understanding and confusion in these interactions. However, an individual's expressions may vary depending on other factors. In cases where these factors are unknown, machine learning methods that infer understanding from nonverbal cues become unreliable. Stress for example has been shown to affect human expression, but it is not clear from the current research how stress affects the expression of understanding.
To address this gap, we design a paradigm that induces understanding and confusion through game rule explanations. During the explanations, self-perceived understanding and confusion are annotated by the participants. A stress condition is also introduced to enable the investigation of changes in the expression of social signals under stress.
We conducted a study to validate the stress induction and participants reported a statistically significant increase in stress during the stress condition compared to the neutral control condition. 
Additionally, feedback from participants shows that the paradigm is effective in inducing understanding and confusion. 
This paradigm paves the way for further studies investigating social signals of understanding to improve human-machine explanation interactions for varying contexts.}},
  author       = {{Paletschek, Jonas}},
  booktitle    = {{12th International Conference on  Affective Computing & Intelligent Interaction}},
  keywords     = {{Understanding, Nonverbal Social Signals, Stress Induction, Explanation, Machine Learning Bias}},
  location     = {{Glasgow}},
  publisher    = {{IEEE}},
  title        = {{{A Paradigm to Investigate Social Signals of Understanding and Their Susceptibility to Stress}}},
  doi          = {{10.1109/ACII63134.2024.00040}},
  year         = {{2024}},
}

@article{61290,
  abstract     = {{ffective computing often relies on audiovisual data to identify affective states from non-verbal signals, such as facial expressions and vocal cues. Since automatic affect recognition can be used in sensitive applications, such as healthcare and education, it is crucial to understand how models arrive at their decisions. Interpretability of machine learning models is the goal of the emerging research area of Explainable AI (explainable AI (XAI)). This scoping review aims to survey the field of audiovisual affective machine learning to identify how XAI is applied in this domain. We first provide an overview of XAI concepts relevant to affective computing. Next, following the recommended PRISMA guidelines, we perform a literature search in the ACM, IEEE, Web of Science and PubMed databases. After systematically reviewing 1190 articles, a final set of 65 papers is included in our analysis. We quantitatively summarize the scope, methods and evaluation of the XAI techniques used in the identified papers. Our findings show encouraging developments for using XAI to explain models in audiovisual affective computing, yet only a limited set of methods are used in the reviewed works. Following a critical discussion, we provide recommendations for incorporating interpretability in future work for affective machine learnin}},
  author       = {{Johnson, David and Hakobyan, Olya and Paletschek, Jonas and Drimalla, Hanna}},
  issn         = {{1949-3045}},
  journal      = {{IEEE Transactions on Affective Computing}},
  number       = {{2}},
  pages        = {{518--536}},
  publisher    = {{Institute of Electrical and Electronics Engineers (IEEE)}},
  title        = {{{Explainable AI for Audio and Visual Affective Computing: A Scoping Review}}},
  doi          = {{10.1109/taffc.2024.3505269}},
  volume       = {{16}},
  year         = {{2024}},
}

@inproceedings{55429,
  abstract     = {{A detailed understanding of the cognitive process underlying diagnostic reasoning in medical experts is currently lacking. While high-level theories like hypothetico-deductive reasoning were proposed long ago, the inner workings of the step-by-step dynamics within the mind remain unknown. We present a fully automated approach to elicit, monitor, and record diagnostic reasoning processes at a fine-grained level. A web-based user interface enables physicians to carry out a full diagnosis process on a simulated patient, given as a pre-defined clinical vignette. By collecting the physician’s information queries and hypothesis revisions, highly detailed diagnostic reasoning trajectories are captured leading to a diagnosis and its justification. Four expert epileptologists with a mean experience of 19 years were recruited to evaluate the system and share their impressions in semi-structured interviews. We find that the recorded trajectories validate proposed theories on broader diagnostic reasoning, while also providing valuable additional details extending previous findings.}},
  author       = {{Battefeld, Dominik and Mues, Sigrid and Wehner, Tim and House, Patrick and Kellinghaus, Christoph and Wellmer, Jörg and Kopp, Stefan}},
  booktitle    = {{Proceedings of the 46th Annual Conference of the Cognitive Science Society}},
  keywords     = {{Differential Diagnosis, Diagnostic Reasoning, Reasoning Process Analysis, Seizure, Epilepsy}},
  location     = {{Rotterdam, NL}},
  title        = {{{Revealing the Dynamics of Medical Diagnostic Reasoning as Step-by-Step Cognitive Process Trajectories}}},
  year         = {{2024}},
}

@inproceedings{48355,
  abstract     = {{Unsupervised speech disentanglement aims at separating fast varying from
slowly varying components of a speech signal. In this contribution, we take a
closer look at the embedding vector representing the slowly varying signal
components, commonly named the speaker embedding vector. We ask, which
properties of a speaker's voice are captured and investigate to which extent do
individual embedding vector components sign responsible for them, using the
concept of Shapley values. Our findings show that certain speaker-specific
acoustic-phonetic properties can be fairly well predicted from the speaker
embedding, while the investigated more abstract voice quality features cannot.}},
  author       = {{Rautenberg, Frederik and Kuhlmann, Michael and Wiechmann, Jana and Seebauer, Fritz and Wagner, Petra and Haeb-Umbach, Reinhold}},
  booktitle    = {{ITG Conference on Speech Communication}},
  location     = {{Aachen}},
  title        = {{{On Feature Importance and Interpretability of Speaker Representations}}},
  year         = {{2023}},
}

@inproceedings{48410,
  author       = {{Wiechmann, Jana and Rautenberg, Frederik and Wagner, Petra and Haeb-Umbach, Reinhold}},
  booktitle    = {{20th International Congress of the Phonetic Sciences (ICPhS) }},
  title        = {{{Explaining voice characteristics to novice voice practitioners-How successful is it?}}},
  year         = {{2023}},
}

@inproceedings{48595,
  author       = {{Peters, Tobias Martin and Visser, Roel W.}},
  booktitle    = {{Communications in Computer and Information Science}},
  isbn         = {{9783031440694}},
  issn         = {{1865-0929}},
  publisher    = {{Springer Nature Switzerland}},
  title        = {{{The Importance of Distrust in AI}}},
  doi          = {{10.1007/978-3-031-44070-0_15}},
  year         = {{2023}},
}

@article{48777,
  abstract     = {{<jats:title>Abstract</jats:title><jats:p>Explainable artificial intelligence has mainly focused on static learning scenarios so far. We are interested in dynamic scenarios where data is sampled progressively, and learning is done in an incremental rather than a batch mode. We seek efficient incremental algorithms for computing feature importance (FI). Permutation feature importance (PFI) is a well-established model-agnostic measure to obtain global FI based on feature marginalization of absent features. We propose an efficient, model-agnostic algorithm called iPFI to estimate this measure incrementally and under dynamic modeling conditions including concept drift. We prove theoretical guarantees on the approximation quality in terms of expectation and variance. To validate our theoretical findings and the efficacy of our approaches in incremental scenarios dealing with streaming data rather than traditional batch settings, we conduct multiple experimental studies on benchmark data with and without concept drift.</jats:p>}},
  author       = {{Fumagalli, Fabian and Muschalik, Maximilian and Hüllermeier, Eyke and Hammer, Barbara}},
  issn         = {{0885-6125}},
  journal      = {{Machine Learning}},
  keywords     = {{Artificial Intelligence, Software}},
  publisher    = {{Springer Science and Business Media LLC}},
  title        = {{{Incremental permutation feature importance (iPFI): towards online explanations on data streams}}},
  doi          = {{10.1007/s10994-023-06385-y}},
  year         = {{2023}},
}

@inproceedings{44853,
  author       = {{Alpsancar, Suzana}},
  booktitle    = {{International Conference on Computer Ethics 2023}},
  location     = {{Chicago, Illinois}},
  number       = {{1}},
  pages        = {{1----17}},
  title        = {{{What is AI Ethics? Ethics as means of self-regulation and the need for critical reflection }}},
  volume       = {{1}},
  year         = {{2023}},
}

@article{51345,
  abstract     = {{<jats:p> The algorithmic imaginary as a theoretical concept has received increasing attention in recent years as it aims at users’ appropriation of algorithmic processes operating in opacity. But the concept originally only starts from the users’ point of view, while the processes on the platforms’ side are largely left out. In contrast, this paper argues that what is true for users is also valid for algorithmic processes and the designers behind. On the one hand, the algorithm imagines users’ future behavior via machine learning, which is supposed to predict all their future actions. On the other hand, the designers anticipate different actions that could potentially performed by users with every new implementation of features such as social media feeds. In order to bring into view this permanently reciprocal interplay coupled to the imaginary, in which not only the users are involved, I will argue for a more comprehensive and theoretically precise algorithmic imaginary referring to the theory of Cornelius Castoriadis. In such a perspective, an important contribution can be formulated for a theory of social media platforms that goes beyond praxeocentrism or structural determinism. </jats:p>}},
  author       = {{Schulz, Christian}},
  issn         = {{0163-4437}},
  journal      = {{Media, Culture & Society}},
  keywords     = {{Sociology and Political Science, Communication}},
  number       = {{3}},
  pages        = {{646--655}},
  publisher    = {{SAGE Publications}},
  title        = {{{A new algorithmic imaginary}}},
  doi          = {{10.1177/01634437221136014}},
  volume       = {{45}},
  year         = {{2023}},
}