LLM-based In-situ Thought Exchanges for Critical Paper ReadingCritical reading is a primary way through which researchers develop their critical thinking skills. While exchanging thoughts and opinions with peers can strengthen critical reading, junior researchers often lack access to peers who can offer diverse perspectives. To address this gap, we designed an in-situ thought exchange interface informed by peer feedback from a formative study (N=8) to support junior researchers’ critical paper reading. We evaluated the effects of thought exchanges under three conditions (no-agent, single-agent, and multi-agent) with 46 junior researchers over two weeks. Our results showed that incorporating agent-mediated thought exchanges during paper reading significantly improved participants’ critical thinking scores compared to the no-agent condition. In the single-agent condition, participants more frequently made reflective annotations on the paper content. In the multi-agent condition, participants engaged more actively with agents’ responses. Our qualitative analysis further revealed that participants compared and analyzed multiple perspectives in the multi-agent condition. This work contributes to understanding in-situ AI-based support for critical paper reading through thought exchanges and offers design implications for future research.2026XFXinrui Fang et al.The University of TokyoHuman-LLM CollaborationUser Research Methods (Interviews, Surveys, Observation)Prototyping & User TestingIUI
A Multi-Factorial Comparative Analysis of Perceived Privacy Violations Caused by Smart Speakers in Germany and the UKSmart speakers pose privacy risks to users and bystanders. We do not know how these risks are perceived depending on different factors, such as the potential privacy violators, the nature of the privacy violation, the different user groups, and culture. Understanding these perceptions is crucial to providing adequate privacy solutions and legislation. To this end, 1,768 participants from Germany and the UK answered our online-questionnaire about their perceptions of five different actors’ possibilities, intentions, and legal bases to commit five privacy violations: data access, data inference, overhearing conversations, secondary use, and passing data along. Participants expressed mild concerns about the main user but greater worry about manufacturers and the state. We observe growing concern among younger people, especially in the UK and that users who do not own the smart speaker are the least concerned group. Our approach can be used to better differentiate perceptions of concerns in other contexts.2025PKPatrick Kühtreiber et al.Privacy by Design & User ControlPrivacy Perception & Decision-MakingIoT Device PrivacyUIST
Beyond Click to Cognition: Effective Interventions for Promoting Examination of False Beliefs in MisinformationIn the digital information ecosystem, clicks serve as a crucial gateway to fact-checking, yet the essential challenge extends beyond this to fostering cognitive shifts that update entrenched false beliefs. This study investigates effective interventions aimed at encouraging users vulnerable to misinformation, particularly those who tend to avoid incongruent facts, to examine their false beliefs. We conducted an online experiment with 627 participants, comparing metacognitive and ranking interventions. Both interventions successfully improved click behavior, with the metacognitive intervention increasing belief-examining clicks by 14 percentage points and the ranking intervention by 33 percentage points. However, only the metacognitive intervention significantly promoted users' examination of misinformation. This finding underscores the importance of interventions that go beyond merely influencing easily measurable clicks to facilitating thoughtful engagement with fact-checking content. We discuss implications for designing strategies to enhance online fact-checking engagement and mitigate misinformation's societal impact.2025YTYuko Tanaka et al.Nagoya Institute of Technology, Graduate School of EngineeringExplainable AI (XAI)Misinformation & Fact-CheckingAlgorithmic Fairness & BiasCHI
Fair Machine Guidance to Enhance Fair Decision Making in Biased PeopleTeaching unbiased decision-making is crucial for addressing biased decision-making in daily life. Although both raising awareness of personal biases and providing guidance on unbiased decision-making are essential, the latter topics remains under-researched. In this study, we developed and evaluated an AI system aimed at educating individuals on making unbiased decisions using fairness-aware machine learning. In a between-subjects experimental design, 99 participants who were prone to bias performed personal assessment tasks. They were divided into two groups: a) those who received AI guidance for fair decision-making before the task and b) those who received no such guidance but were informed of their biases. The results suggest that although several participants doubted the fairness of the AI system, fair machine guidance prompted them to reassess their views regarding fairness, reflect on their biases, and modify their decision-making criteria. Our findings provide insights into the design of AI systems for guiding fair decision-making in humans.2024MYMingzhe Yang et al.The University of TokyoAI-Assisted Decision-Making & AutomationAI Ethics, Fairness & AccountabilityCHI
PATCH: A Plug-in Framework of Non-blocking Inference for Distributed Multimodal System"Recent advancements in deep learning have shown that multimodal inference can be particularly useful in tasks like autonomous driving, human health, and production line monitoring. However, deploying state-of-the-art multimodal models in distributed IoT systems poses unique challenges since the sensor data from low-cost edge devices can get corrupted, lost, or delayed before reaching the cloud. These problems are magnified in the presence of asymmetric data generation rates from different sensor modalities, wireless network dynamics, or unpredictable sensor behavior, leading to either increased latency or degradation in inference accuracy, which could affect the normal operation of the system with severe consequences like human injury or car accident. In this paper, we propose PATCH, a framework of speculative inference to adapt to these complex scenarios. PATCH serves as a plug-in module in the existing multimodal models, and it enables speculative inference of these off-the-shelf deep learning models. PATCH consists of 1) a Masked-AutoEncoder-based cross-modality imputation module to impute missing data using partially-available sensor data, 2) a lightweight feature pair ranking module that effectively limits the searching space for the optimal imputation configuration with low computation overhead, and 3) a data alignment module that aligns multimodal heterogeneous data streams without using accurate timestamp or external synchronization mechanisms. We implement PATCH in nine popular multimodal models using five public datasets and one self-collected dataset. The experimental results show that PATCH achieves up to 13% mean accuracy improvement over the state-of-art method while only using 10% of training data and reducing the training overhead by 73% compared to the original cost of retraining the model." https://doi.org/10.1145/36108852023JWJuexing Wang et al.Teleoperated DrivingV2X (Vehicle-to-Everything) Communication DesignContext-Aware ComputingUbiComp
Use of an AI-powered Rewriting Support Software in Context with Other Tools: A Study of Non-Native English SpeakersAcademic writing in English can be challenging for non-native English speakers (NNESs). AI-powered rewriting tools can potentially improve NNESs' writing outcomes at a low cost. However, whether and how NNESs make valid assessments of the revisions provided by these algorithmic recommendations remains unclear. We report a study where NNESs leverage an AI-powered rewriting tool, Langsmith, to polish their drafted academic essays. We examined the participants' interactions with the tool via user studies and interviews. Our data reveal that most participants used Langsmith in combination with other tools, such as machine translation (MT), and those who used MT had different ways of understanding and evaluating Langsmith's suggestions than those who did not. Based on these findings, we assert that NNESs' quality assessment in AI-powered rewriting tools is influenced by the simultaneous use of multiple tools, offering valuable insights into the design of future rewriting tools for NNESs.2023TITakumi Ito et al.Generative AI (Text, Image, Music, Video)AI-Assisted Creative WritingUIST
Who Does Not Benefit from Fact-checking Websites? A Psychological Characteristic Predicts the Selective Avoidance of Clicking Uncongenial FactsFact-checking messages are shared or ignored subjectively. Users tend to seek like-minded information and ignore information that conflicts with their preexisting beliefs, leaving like-minded misinformation uncontrolled on the Internet. To understand the factors that distract fact-checking engagement, we investigated the psychological characteristics associated with users’ selective avoidance of clicking uncongenial facts. In a pre-registered experiment, we measured participants’ (N = 506) preexisting beliefs about COVID-19-related news stimuli. We then examined whether they clicked on fact-checking links to false news that they believed to be accurate. We proposed an index that divided participants into fact-avoidance and fact-exposure groups using a mathematical baseline. The results indicated that 43% of participants selectively avoided clicking on uncongenial facts, keeping 93% of their false beliefs intact. Reflexiveness is the psychological characteristic that predicts selective avoidance. We discuss susceptibility to click bias that prevents users from utilizing fact-checking websites and the implications for future design.2023YTYuko Tanaka et al.Nagoya Institute of TechnologyVoice AccessibilityMisinformation & Fact-CheckingCHI
Preserving Agency During Electrical Muscle Stimulation Training Speeds up Reaction Time Directly After Removing EMSForce feedback devices, such as motor-based exoskeletons or wearables based on electrical muscle stimulation (EMS), have the unique potential to accelerate users’ own reaction time (RT). However, this speedup has only been explored while the device is attached to the user. In fact, very little is known regarding whether this faster reaction time still occurs after the user removes the device from their bodies–this is precisely what we investigated by means of a simple reaction time (RT) experiment, in which participants were asked to tap as soon as they saw an LED flashing. Participants experienced this in three EMS conditions: (1) fast-EMS, the electrical impulses were synced with the LED; (2) agency-EMS, the electrical impulse was delivered 40ms faster than the participant’s own RT, which prior work has shown to preserve one’s sense of agency over this movement; and, (3) late-EMS: the impulse was delivered after the participant’s own RT. Our results revealed that the participants’ RT was significantly reduced by approximately 8ms(up to 20ms) only after training with the agency-EMS condition. This finding suggests that the prioritizing agency during EMS training is key to motor-adaptation, i.e., it enables a faster motor response even after the user has removed the EMS device from their body.2021SKShunichi Kasahara et al.Sony CSL, The University of TokyoVibrotactile Feedback & Skin StimulationElectrical Muscle Stimulation (EMS)CHI