It is crucial that ASR systems can handle the wide range of variations in speech of speakers from different demographic groups, with different speaking styles, and of speakers with (dis)abilities. A potential quality-of-service harm arises when ASR systems do not perform equally well for everyone. ASR systems may exhibit bias against certain types of speech, such as non-native accents, different age groups and gender. In this study, we evaluate two widely-used neural network-based architectures: Wav2vec2 and Whisper on potential biases for Dutch speakers. We used the Dutch speech corpus JASMIN as a test set containing read and conversational speech in a human-machine interaction setting. The results reveal a significant bias against non-natives, children and elderly and some regional dialects. The ASR systems generally perform slightly better for women than for men.
MULTIFILE
Albeit the widespread application of recommender systems (RecSys) in our daily lives, rather limited research has been done on quantifying unfairness and biases present in such systems. Prior work largely focuses on determining whether a RecSys is discriminating or not but does not compute the amount of bias present in these systems. Biased recommendations may lead to decisions that can potentially have adverse effects on individuals, sensitive user groups, and society. Hence, it is important to quantify these biases for fair and safe commercial applications of these systems. This paper focuses on quantifying popularity bias that stems directly from the output of RecSys models, leading to over recommendation of popular items that are likely to be misaligned with user preferences. Four metrics to quantify popularity bias in RescSys over time in dynamic setting across different sensitive user groups have been proposed. These metrics have been demonstrated for four collaborative filteri ng based RecSys algorithms trained on two commonly used benchmark datasets in the literature. Results obtained show that the metrics proposed provide a comprehensive understanding of growing disparities in treatment between sensitive groups over time when used conjointly.
DOCUMENT
Valuation judgement bias has been a research topic for several years due to its proclaimed effect on valuation accuracy. However, little is known on the emphasis of literature on judgement bias, with regard to, for instance, research methodologies, research context and robustness of research evidence. A synthesis of available research will establish consistency in the current knowledge base on valuer judgement, identify future research opportunities and support decision-making policy by educational and regulatory stakeholders how to cope with judgement bias. This article therefore, provides a systematic review of empirical research on real estate valuer judgement over the last 30 years. Based on a number of inclusion and exclusion criteria, we have systematically analysed 32 relevant papers on valuation judgement bias. Although we find some consistency in evidence, we also find the underlying research to be biased; the methodology adopted is dominated by a quantitative approach; research context is skewed by timing and origination; and research evidence seems fragmented and needs replication. In order to obtain a deeper understanding of valuation judgement processes and thus extend the current knowledge base, we advocate more use of qualitative research methods and scholars to adopt an interpretative paradigm when studying judgement behaviour.
DOCUMENT
This article interrogates platform-specific bias in the contemporary algorithmic media landscape through a comparative study of the representation of pregnancy on the Web and social media. Online visual materials such as social media content related to pregnancy are not void of bias, nor are they very diverse. The case study is a cross-platform analysis of social media imagery for the topic of pregnancy, through which distinct visual platform vernaculars emerge. The authors describe two visualization methods that can support comparative analysis of such visual vernaculars: the image grid and the composite image. While platform-specific perspectives range from lists of pregnancy tips on Pinterest to pregnancy information and social support systems on Twitter, and pregnancy humour on Reddit, each of the platforms presents a predominantly White, able-bodied and heteronormative perspective on pregnancy.
DOCUMENT
The study evaluated two speech recognition systems, Wav2vec2 and Whisper, for potential biases for Dutch speakers.Results obtained by evaluating on the JASMIN corpus revealed biases against non-native speakers, children, and the elderly,with (slightly) better performance for women. The study emphasizes the need for ASR systems to handle variations in speakingin order to reach equal performance among all users.
DOCUMENT
In this paper, we report on the initial results of an explorative study that aims to investigate the occurrence of cognitive biases when designers use generative AI in the ideation phase of a creative design process. When observing current AI models utilised as creative design tools, potential negative impacts on creativity can be identified, namely deepening already existing cognitive biases but also introducing new ones that might not have been present before. Within our study, we analysed the emergence of several cognitive biases and the possible appearance of a negative synergy when designers use generative AI tools in a creative ideation process. Additionally, we identified a new potential bias that emerges from interacting with AI tools, namely prompt bias.
DOCUMENT
Narrative structures such as the Hero’s Journey and Heroine’s Journey have long influenced how characters, themes, and roles are portrayed in storytelling. When used to guide narrative generation in systems powered by Large Language Models (LLMs), these structures may interact with model-internal biases, reinforcing traditional gender norms. This workshop examines how protagonist gender and narrative structure shape storytelling outcomes in LLM-based storytelling systems. Through hands-on experiments and guided analysis, participants will explore gender representation in LLM-generated stories, perform counterfactual modifications, and evaluate how narrative interpretations shift when character gender is altered. The workshop aims to foster interdisciplinary collaborations, inspire novel methodologies, and advance research on fair and inclusive AI-driven storytelling in games and interactive media.
LINK
Traditional IMU based PDR systems suffer from rapidly growing drift effects due to the inherent bias of the inertial sensor. Many existing solutions to mitigate this problem use aiding sensors or information as heuristics or map data. We propose a new optimization framework to solve the PDR estimation problem where the sensors biases are explicitly included as state variables and therefore be used to correct for bias effects in the PDR. By using a smoothing approach and exploiting the rigid structure of a MIMU array one can solve for the slowly varying sensor biases. This paper presents the method and gives an exemplary result of a walking trial. Good agreements in the position and orientation with an optical reference system were found. Moreover, accelerometer and gyroscope biases could be estimated accordingly. Further research includes the performance of more experiments under various conditions such that a more quantitative evaluation can be obtained. In addition, an exploration of a (pseudo) realtime filter version would be valuable such that the system can be applied online.
MULTIFILE
Purpose – Information verification is an important factor in commercial valuation practice.Valuers use their professional autonomy to decide on the level of verification required, thereby creating an opportunity for client-related judgement bias in valuation. The purpose of this paper is to assess the manifestation of client attachment risks in information verification. Design/methodology/approach – A case-based questionnaire was used to retrieve data from 290 commercial valuation professionals in the Netherlands, providing a 15 per cent response rate of the Dutch commercial valuation population. Descriptive and inferential statistics have been used to test research hypotheses involving relations between information verification and professional features that may indicate client attachment such as an executive job level and brokerage experience. Findings – The results reveal that valuers acting at partner level within their organisation obtain lower scores on information verification compared to lower-ranked valuers. Also, brokerage experience correlates negatively to information verification of valuation professionals. Both findings have statistical significance. Research limitations/implications – The results reflect valuers’ reasoning behaviour rather than actual behaviour. Replication of findings through experimental design will contribute to research validity. Practical implications – Maintaining close client contact in a competitive environment is important for business continuity yet may foster client attachment.The associated downside risks in valuation practice call for higher awareness of (subconscious) client influence and the development of attitudinal scepticism in valuer training programmes. Originality/value – This paper is one of the few that explore possible sources of valuer judgement bias by relating client-friendly valuer features to a key area of valuation i.e. information verification.
LINK
This thesis provides an examination of judgement autonomy of Dutch commercial real estate valuers in relation to client orientation. The valuation of commercial real estate such as offices or retail properties requires in-depth analysis due to its uniqueness by location, building type and usage details. Essentially, a register-valuer is qualified and instructed to assess a property value to one’s best cognitive effort and inform others of this outcome by means of a valuation report. In the Netherlands, concerns over independence risks and client-related judgement risks of valuers have been raised by regulative authorities as the Dutch Central Bank (DNB) and the Dutch Authority for the Financial Markets (AFM). A significant part of these concerns followed the 2008 financial crisis, which appeared to be at least partially driven by unreliable and incomparable valuations of Dutch commercial real estate (AFM, 2014; DNB, 2012; 2015). Among other things, these concerns led to the instigation of the Nederlands Register Vastgoed Taxateurs (NRVT) in 2015. NRVT is a new Dutch central register of valuation practitioners set up in order to improve self-regulation, quality control and compliance of valuation practitioners. Currently, the chamber for commercial real estate valuation holds about 2,000 commercial valuation registrations (NRVT, 2020). The introduction of NRVT, and other measures taken, reflect an instrumental view towards enhancing professionalism of Dutch valuers. This view is based on a systematic orientation to professional conduct in which good practice is primarily objectively determined (Van Ewijk, 2019). However, Wassink and Bakker (2016) point out that individuals make personal choices in order to deal with work complexity. Insight into and reflection on individual choices is part of what is referred to as normative aspects of professionalisation: what norms prevail in individual judgement and decision-making and why (Van Ewijk, 2019). In this regard, insight into judgement reasoning of valuation practitioners may contribute to normative levels of professional development of valuers. The need for such is expressed through community concerns over how individual judgement autonomy may become subdued due to instrumental-driven developments taking place in the sector. The combination of authoritative concerns over professional quality in the Netherlands and lack of (scientific) insight on how client influence affects judgement in valuation practice poses a problem: How may practitioners address client-related judgement bias risks and improve valuation accuracy from this viewpoint, if little is known on how such risks may occur in daily practice? The seemingly scarce scientific insights available in this regard in the Netherlands may also prevent educational programs to adequately address valuer independence and objectivity risks in relevant training programs. In order to address this knowledge gap, the present PhD research examines the following research problem: 169 Summary “How does client orientation affect professional judgement autonomy of commercial real estate valuers in the Netherlands?” The term ‘client orientation’ should be broadly interpreted and may refer to valuers’ perception, understanding and meaning given to alleged, actual or anticipated client-related aspects. Information on such client aspects is not required for the performance of valuation instructions. It should also be noted that this research examines the context of how client orientation may affect valuer judgement reasoning patterns during work practice, yet not its effect in terms of decision on final value opinion.
MULTIFILE
