28 November 2024 in International Journal of Human–Computer Interaction
8 November 2021 in Applied AI Letters
Explainable artificial intelligence (XAI) systems aim to provide users with information to help them better understand computational models and reason about why outputs were generated. However, there are many different ways an XAI interface might present explanations, which makes designing an appropriate and effective interface an important and challenging task. Our work investigates how different types and amounts of explanatory information affect user ability to utilize explanations to understand system behavior and improve task performance. The presented research employs a system for detecting the truthfulness of news statements. In a controlled experiment, participants were tasked with using the system to assess news statements as well as to learn to predict the output of the AI. Our experiment compares various levels of explanatory information to contribute empirical data about how explanation detail can influence utility. The results show that more explanation information improves participant understanding of AI models, but the benefits come at the cost of time and attention needed to make sense of the explanation.
1 January 2009 in Methods of Information in Medicine
Summary: Objectives: Recently, ontological principles have been applied to numerous biomedical vocabularies, with the intention to identify mistakes and poor modeling decisions. No doubt, such applications are useful and necessary for terminological systems like SNOMED CT based on an axiomatic logical formalism. Methods: In the following review, ontology is dealt with by focussing on particularly two aspects: the problem of ISA-overloading and the intrusion of epistemology-loaded terms in biomedical vocabularies. Both aspects are considered with respect to three types of biomedical vocabularies. Results: Opposed to concept-oriented terminological systems, the purpose-specific organization of descriptors in thesauri and classes in statistical classifications on an extra aggregation level make it impossible to apply ontological principles. On the contrary, their intended purpose presupposes specific mechanisms that are in conflict with those principles. Conclusions: Interestingly, for thesauri and classifications there are rather similar initiatives linking the extra level of descriptors and classes on the one hand and an intermediate concept level on the other hand. Such an approach proved beneficial for maintaining and translating thesauri and classifications.
25 July 2022 in IEEE Transactions on Visualization and Computer Graphics
Information visualization uses various types of representations to encode data into graphical formats. Prior work on visualization techniques has evaluated the accuracy of perceived numerical data values from visual data encodings such as graphical position, length, orientation, size, and color. Our work aims to extend the research of graphical perception to the use of motion as data encodings for quantitative values. We present two experiments implementing multiple fundamental aspects of motion such as type, speed, and synchronicity that can be used for numerical value encoding as well as comparing motion to static visual encodings in terms of user perception and accuracy. We studied how well users can assess the differences between several types of motion and static visual encodings and present an updated ranking of accuracy for quantitative judgments. Our results indicate that non-synchronized motion can be interpreted more quickly and more accurately than synchronized motion. Moreover, our ranking of static and motion visual representations shows that motion, especially expansion and translational types, has great potential as a data encoding technique for quantitative value. Finally, we discuss the implications for the use of animation and motion for numerical representations in data visualization.
14 October 2024 in Proceedings of the AAAI Conference on Human Computation and Crowdsourcing
Crowdsourcing platforms have traditionally been designed with a focus on workstation interfaces, restricting the flexibility that crowdworkers need. Recognizing this limitation and the need for more adaptable platforms, prior research has highlighted the diverse work processes of crowdworkers, influenced by factors such as device type and work stage. However, these variables have largely been studied in isolation. Our study is the first to explore the interconnected variabilities among these factors within the crowdwork community. Through a survey involving 150 Amazon Mechanical Turk crowdworkers, we uncovered three distinct groups characterized by their interrelated variabilities in key work aspects. The largest group exhibits a reliance on traditional devices, showing limited interest in integrating smartphones and tablets into their work routines. The second-largest group also primarily uses traditional devices but expresses a desire for supportive tools and scripts that enhance productivity across all devices, particularly smartphones and tablets. The smallest group actively uses and strongly prefers non-workstation devices, especially smartphones and tablets, for their crowdworking activities. We translate our findings into design insights for platform developers, discussing the implications for creating more personalized, flexible, and efficient crowdsourcing environments. Additionally, we highlight the unique work practices of these crowdworker clusters, offering a contrast to those of more traditional and established worker groups.