While large language model (LLM)-based user profiling offers significant potential for personalization, most existing approaches rely on empirical heuristics and lack grounding in the psychological mechanism that drive human behavior. In this paper, we introduce TRIPLE (Theory-guided Reasoning for Intent and habIt Profiling with LLMs for pErsonalization), a novel framework that systematically integrates dual-process theory from social psychology into LLM-based user modeling. TRIPLE (1) constructs a habitual behavior profile by identifying repeated patterns over time to model automatic responses; (2) builds an intentional behavior profile by inferring user attitudes, subjective norms and perceived behavioral control based on the Theory of Planned Behavior (TPB); and (3) generates behavioral rationale that reveal the interaction between habitual and intentional processes to predict user behavior in context-specific situations. We evaluate TRIPLE on five personalization tasks from the LaMP benchmark using multiple open-source LLMs. Results show that TRIPLE consistently outperforms existing in-context learning methods, with especially pronounced gains on complex generative tasks such as headline and title generation. Qualitative analyses further demonstrate that the profiles and reasoning paths generated by TRIPLE provide interpretable and psychologically grounded explanations of user behavior. These findings provide strong evidence that incorporating validated behavioral theories into LLM-based personalization enhances both predictive performance and interpretability, paving a way for theory-driven, socio-cognitively informed user modeling.

Public surveys are indispensable resources for understanding social dynamics, yet their analysis often imposes a high cognitive load due to structural complexity. In this paper, we present LAPS, a Large Language Model (LLM)-assisted automated framework that supports end-to-end, hypothesis-driven statistical analysis of survey data. LAPS consists of four modules (i.e., Operationalization, Planning, Execution, and Reporting) with human-in-the-loop mechanisms to balance automation with user agency. To evaluate the applicability of LAPS, we conducted a within-subjects user study with 12 social science researchers across three analytical environments: traditional statistical tools, a general-purpose LLM, and LAPS. Our findings demonstrate that LAPS ensures researcher agency and analytical stability, reduces the cognitive burden in the analysis workflow, and produces trustworthy, coherent outputs. Based on these findings, we reflect on how LAPS improves researchers’ workflows and discuss design implications for scalable and trustworthy human-AI collaboration in survey-based research.

Designers often regard vagueness as an essential aspect of creative work, as it fosters diverse interpretations and helps prevent fixation. Although large language models (LLMs) are increasingly viewed as a promising creative partner, designers struggle to productively incorporate vagueness into AI-supported workflows. To address this challenge, we present QuerySwitch, an interactive prototype that enables fashion designers to manage vagueness by flexibly switching between two distinct query-output modes. Findings from a user study show that QuerySwitch helps fashion designers balance vagueness, enhances the usability of LLMs in design tasks, and promotes creative exploration. This work contributes to HCI by (1) foregrounding a critical construct in human–AI collaboration, (2) demonstrating how interaction mechanisms can scaffold designer agency in LLMs use, and (3) articulating design principles—structuring exploration and preserving key query formulations—that extend to creativity-driven domains.

Beyond hallucinations, Large Language Models (LLMs) can craft deceptive arguments that erode users’ critical thinking, posing a significant yet underexamined societal risk. To address this gap, we develop a taxonomy of eight deceptive persuasion strategies by integrating top-down rhetorical theory with a bottom-up analysis of 3,360 AI-generated messages by four LLM families and examining their effects on user perceptions. Through a large-scale user study (N=602) complemented by a think-aloud protocol, we found that participants were vulnerable to Information Manipulation and Uncertainty Exploitation, especially when a message contradicted their prior beliefs. Vulnerability was significantly higher for participants with low cognitive reflection, low topic knowledge, and low topic involvement. Qualitative analyses further revealed that participants were persuaded by the plausibility of an overall narrative even when they distrusted specific details, interpreting deceptive outputs as logically framed information that broadens perspective. We discuss critical implications of these findings for the design of trustworthy AI systems, adaptive user interfaces, and targeted literacy education.

As personality can be useful in many cases, such as better understanding people’s underlying contexts or providing personalized services, research has long focused on modeling personality from data. However, the development of personality detection models faces challenges due to the inherent latent and relative characteristics of personality, as well as the lack of annotated datasets. To address these challenges, our research focuses on methods that effectively exploit the inherent knowledge of Large Language Models (LLMs). We propose a novel approach that compares contrasting perspectives to better capture the relative nature of personality traits. In this paper, we introduce PADO (Personality-induced multi-Agent framework for Detecting OCEAN of the Big Five personality traits), the first LLM-based multi-agent personality detection framework. PADO employs personality-induced agents to analyze text from multiple perspectives, followed by a comparative judgment process to determine personality trait levels. Our experiments with various LLM models, from GPT-4o to LLaMA3-8B, demonstrate PADO’s effectiveness and generalizability, especially with smaller parameter models. This approach offers a more nuanced, context-aware method for personality detection, potentially improving personalized services and insights into digital behavior. We will release our codes.

Vision transformers use [CLS] token to predict image classes. Their explainability visualization has been studied using relevant information from the [CLS] token or focusing on attention scores during self-attention. However, such visualization is challenging because of the dependence of the interpretability of a vision transformer on skip connections and attention operators, the instability of non-linearities in the learning process, and the limited reflection of self-attention scores on relevance. We argue that the output patch embeddings in a vision transformer preserve the image information of each patch location, which can facilitate the prediction of an image class. In this paper, we propose ICEv2 (ICEv2: $${{{\underline{\varvec{I}}}}}$$ nterpretability, $${{{\underline{\varvec{C}}}}}$$ omprehensiveness, and $${{{\underline{\varvec{E}}}}}$$ xplainability in Vision Transformer), an explainability visualization method that addresses the limitations of ICE (i.e., high dependence of hyperparameters on performance and the inability to preserve the model’s properties) by minimizing the number of training encoder layers, redesigning the MLP layer, and optimizing hyperparameters along with various model size. Overall, ICEv2 shows higher efficiency, performance, robustness, and scalability than ICE. On the ImageNet-Segmentation dataset, ICEv2 outperformed all explainability visualization methods in all cases depending on the model size. On the Pascal VOC dataset, ICEv2 outperformed both self-supervised and supervised methods on Jaccard similarity. In the unsupervised single object discovery, where untrained classes are present in the images, ICEv2 effectively distinguished between foreground and background, showing performance comparable to the previous state-of-the-art. Lastly, ICEv2 can be trained with significantly lower training computational complexity.

As the perspective on creativity shifts to “how it is expressed,” research has aimed to categorize it by problem-solving style. Since harnessing individual creative traits can positively impact creative performance, there has been an emphasis on designing computer systems that are tailored to personal problem-solving behaviors. AI-CST has opened up the potential to facilitate such customization. In this work, we consider two types of creativity—adaptors and innovators—based on problem-solving styles, and investigate AI-CST designs that both types could flexibly use according to the fashion design process. We identified two main AI-CST functions—determining design direction and receiving design inspiration—of the fashion design process, and developed CoCoStyle to map these functions. Through a user study with 30 fashion professionals (15 adaptors and 15 innovators), we found significant differences between the two groups from survey responses, system usage logs, and interviews. Based on the results, we discuss the theoretical and practical implications of AI and AI-CST where creativity is essential.

Maintaining a daily routine has profound implications for physical, emotional, and social well-being. Autistic individuals may experience various challenges in establishing and maintaining a healthy daily routine due to their tendency to be inactive in daily life combined with their characteristics and preferences. Previous studies employing mobile technology to support autistic individuals have primarily focused on self-help functions, with limited exploration into the detailed needs of these individuals to develop and maintain personalized routines. In this study, we conducted a nine-week field study with 18 autistic individuals using RoutineAid, a gamified app designed to support key routines of autistic individuals (i.e., physical activity, diet, mindfulness, and sleep). Our analysis incorporated five measures of self-evaluation on daily life, app usage logs, Fitbit physical activity data, and interviews. Our findings demonstrate the effectiveness of RoutineAid and highlight its two primary affordances for autistic individuals: (1) promoting self-efficacy and embedding health behavior and (2) refining daily routines for healthier outcomes. We discuss salient design insights for developing daily routine management systems for autistic individuals.

While a few public benchmarks have been proposed for training hate speech detection models, the differences in labeling criteria between these benchmarks pose challenges for generalized learning, limiting the applicability of the models.Previous research has presented methods to generalize models through data integration or augmentation, but overcoming the differences in labeling criteria between datasets remains a limitation.To address these challenges, we propose PREDICT, a novel framework that uses the notion of multi-agent for hate speech detection.PREDICT consists of two phases: (1) PRE (Perspectivebased REasoning): Multiple agents are created based on the induced labeling criteria of given datasets, and each agent generates stances and reasons; (2) DICT (Debate using InCongruenT references): Agents representing hate and nonhate stances conduct the debate, and a judge agent classifies hate or non-hate and provides a balanced reason.Experiments on five representative public benchmarks show that PREDICT achieves superior cross-evaluation performance compared to methods that focus on specific labeling criteria or majority voting methods.Furthermore, we validate that PREDICT effectively mediates differences between agents' opinions and appropriately incorporates minority opinions to reach a consensus.

Considerable efforts are currently underway to mitigate the negative impacts of echo chambers, such as increased susceptibility to fake news and resistance towards accepting scientific evidence. Prior research has presented the development of computer systems that support the consumption of news information from diverse political perspectives to mitigate the echo chamber effect. However, existing studies still lack the ability to effectively support the key processes of news information consumption and quantitatively identify a political stance towards the information. In this paper, we present HearHere, an AI-based web system designed to help users accommodate information and opinions from diverse perspectives. HearHere facilitates the key processes of news information consumption through two visualizations. Visualization 1 provides political news with quantitative political stance information, derived from our graph-based political classification model, and users can experience diverse perspectives (Hear). Visualization 2 allows users to express their opinions on specific political issues in a comment form and observe the position of their own opinions relative to pro-liberal and pro-conservative comments presented on a map interface (Here). Through a user study with 94 participants, we demonstrate the feasibility of HearHere in supporting the consumption of information from various perspectives. Our findings highlight the importance of providing political stance information and quantifying users' political status as a means to mitigate political polarization. In addition, we propose design implications for system development, including the consideration of demographics such as political interest and providing users with initiatives.

Social anxiety disorder is a psychological condition characterized by excessive nervousness in social situations, such as interpersonal interactions. Exposure therapy has shown benefits in its treatment, and virtual reality (VR) technology has gained much attention for reducing physical and psychological distance and providing additional quantitative evidence from the data generated by standard VR devices (e.g., head-mounted display). Clinical psychology studies have highlighted the importance of mediating variables of social anxiety; however, existing VR-based social anxiety studies have neglected such variables with respect to user experience and data analysis in the context of VR, although these variables could provide insights into the design and use of VR for the treatment of social anxiety. In this study, we focused on two representative mediating variables of social anxiety: (1) the gap between self-presentation motivation and expectancy , and (2) self-focused attention . We used sensor data (e.g., head movement, eye movement, eye gaze, and psychological signals) to investigate the impact of these variables on users' anxiety responses in VR. We developed VR-based Social Anxiety Support Tool (VRST) that reflects the theoretical design elements of effective anxiety provocation. Based on the results of a user study with 30 participants, we confirmed that the mediating variables were associated with social anxiety in the VR environment. We also found that the mediating variables were associated with eye gaze, eye pupil, head movement, and body temperature. Our study results provide researchers, designers, and practitioners with empirical evidence and implications for the use of VR technology and sensor data in the mental health context.

Large language model (LLM)-based AI for code model (e.g., Copilot) demonstrates the potential of using AI in specialized domains such as software engineering. While previous research has focused on fine-tuning models with additional data and computational cost to construct models optimized for specific domains, our research focuses on prompt engineering methods that maximize the performance of existing models. We conducted a quantitative and qualitative user study using the AI for code model and identified two limitations that hinder the recommendation performance of the model. We propose two methods to address these limitations through effective prompt engineering. Finally, we identified the potential for the use of our proposed methods to be utilized and discussed the direction of future research for the effective use of the LLM.

Implementing structure into our daily lives is critical for maintaining health, productivity, and social and emotional well-being. New norms for routine management have emerged during the current pandemic, and in particular, individuals with autism find it difficult to adapt to those norms. While much research has focused on the use of computer technology to support individuals with autism, little is known about ways of helping them establish and maintain “self-directed” routine structures. In this paper, we identify design requirements for an app that support four key routine components (i.e., physical activity, diet, mindfulness, and sleep) through a formative study and develop RoutineAid, a gamified smartphone app that reflects the design requirements. The results of a two-month field study on design feasibility highlight two affordances of RoutineAid—the establishment of daily routines by facilitating micro-planning and the maintenance of daily routines through celebratory interactions. We discuss salient design considerations for the future development of daily routine management tools for individuals with autism.

Vision transformers use [CLS] tokens to predict image classes. Their explainability visualization has been studied using relevant information from [CLS] tokens or focusing on attention scores during self-attention. Such visualization, however, is challenging because of the dependence of the structure of a vision transformer on skip connections and attention operators, the instability of non-linearities in the learning process, and the limited reflection of self-attention scores on relevance. We argue that the output vectors for each input patch token in a vision transformer retain the image information of each patch location, which can facilitate the prediction of an image class. In this paper, we propose ICE (Adversarial Normalization: I Can visualize Everything), a novel method that enables a model to directly predict a class for each patch in an image; thus, advancing the effective visualization of the explainability of a vision transformer. Our method distinguishes background from foreground regions by predicting background classes for patches that do not determine image classes. We used the DeiT-S model, the most representative model employed in studies, on the explainability visualization of vision transformers. On the ImageNet-Segmentation dataset, ICE outperformed all explainability visualization methods for four cases depending on the model size. We also conducted quantitative and qualitative analyses on the tasks of weakly-supervised object localization and unsupervised object discovery. On the CUB-200-2011 and PASCALVOC07/12 datasets, ICE achieved comparable performance to the state-of-the-art methods. We incorporated ICE into the encoder of DeiT-S and improved efficiency by 44.01% on the ImageNet dataset over that achieved by the original DeiT-S model. We showed performance on the accuracy and efficiency comparable to EViT, the state-of-the-art pruning model, demonstrating the effectiveness of ICE. The code is available at https://github.com/Hanyang-HCC-Lab/ICE.

As the number of TV shows increases, designing recommendation systems to provide users with their favorable TV shows becomes more important. In a TV show domain, watching a TV show (i.e., giving implicit feedback to the show) among the TV shows broadcast at the same time frame implies that the currently watching show is the winner in the competition with others (i.e., losers). However, in previous studies, such a notion of limited competitions has not been considered in estimating the user’s preferences for TV shows. In this paper, we propose a new recommendation framework to take this new notion into account based on pair-wise models. Our framework is composed of the following ideas: (i) identify winners and losers by determining pairs of competing TV shows; (ii) learn the pairs of competing TV shows based on the confidence for the pair-wise preference between the winner and the loser; (iii) recommend the most favorable TV shows by considering the time factors with respect to users and TV shows. Using a real-world TV show dataset, our experimental results show that our proposed framework consistently improves the accuracy of recommendation by up to 38%, compared with the best state-of-the-art method. The code and datasets of our framework are available in an external link (https://github.com/hongkyun-bae/tvshow_rs).

Virtual reality (VR) has become a valuable tool for social and educational purposes for autistic people, as it provides flexible environmental support to create a variety of experiences. A growing body of recent research has examined the behaviors of autistic people using sensor-based data to better understand autistic people and investigate the effectiveness of VR. Comprehensive analysis of the various signals that can be easily collected in the VR environment can promote understanding of autistic people. While this quantitative evidence has the potential to help both autistic people and others (e.g., autism experts) to understand behaviors of autistic people, existing studies have focused on single signal analysis and have not determined the acceptability of signal analysis results from the autistic person’s point of view. To facilitate the use of multiple sensor signals in VR for autistic people and experts, we introduce V-DAT (Virtual Reality Data Analysis Tool), designed to support a VR sensor data handling pipeline. V-DAT takes into account four sensor modalities—head position and rotation, eye movement, audio, and physiological signals—that are actively used in current VR research for autistic people. We explain the characteristics and processing methods of the data for each modality as well as the analysis with comprehensive visualizations of V-DAT. We also conduct a case study to investigate the feasibility of V-DAT as a way of broadening understanding of autistic people from the perspectives of both autistic people and autism experts. Finally, we discuss issues with the process of V-DAT development and complementary measures for the applicability and scalability of a sensor data management system for autistic people.

The political stance prediction for news articles has been widely studied to mitigate the echo chamber effect – people fall into their thoughts and reinforce their pre-existing beliefs. The previous works for the political stance problem focus on (1) identifying political factors that could reflect the political stance of a news article and (2) capturing those factors effectively. Despite their empirical successes, they are not sufficiently justified in terms of how effective their identified factors are in the political stance prediction. Motivated by this, in this work, we conduct a user study to investigate important factors in political stance prediction, and observe that the context and tone of a news article (implicit) and external knowledge for real-world entities appearing in the article (explicit) are important in determining its political stance. Based on this observation, we propose a novel knowledge-aware approach to political stance prediction (KHAN), employing (1) hierarchical attention networks (HAN) to learn the relationships among words and sentences in three different levels and (2) knowledge encoding (KE) to incorporate external knowledge for real-world entities into the process of political stance prediction. Also, to take into account the subtle and important difference between opposite political stances, we build two independent political knowledge graphs (KG) (i.e., KG-lib and KG-con) by ourselves and learn to fuse the different political knowledge. Through extensive evaluations on three real-world datasets, we demonstrate the superiority of KHAN in terms of (1) accuracy, (2) efficiency, and (3) effectiveness.

In this article, we propose a framework for embedding-based community detection on signed networks, namelyAdversarial learning ofBalanced triangle forCommunity detection, in short${{\sf ABC}}$. It first represents all the nodes of a signed network as vectors in low-dimensional embedding space and conducts a clustering algorithm (e.g.,k-means) on vectors, thereby detecting a community structure in the network. When performing the embedding process,${{\sf ABC}}$learns only the edges belonging to balanced triangles whose edge signs follow the balance theory, significantly excluding noise edges in learning. To address the sparsity of balanced triangles in a signed network,${{\sf ABC}}$learns not only the edges in balancedreal-triangles but those in balancedvirtual-triangles that do not actually exist but are produced by our generator. Finally,${{\sf ABC}}$employs adversarial learning to generate more-realistic balanced virtual-triangles with less noise edges. Through extensive experiments using seven real-world networks, we validate the effectiveness of (1) learning edges belonging to balanced real/virtual-triangles and (2) employing adversarial learning for signed network embedding. We show that${{\sf ABC}}$consistently and significantly outperforms the state-of-the-art community detection methods in all datasets.

Reflecting on stress-related data is critical in addressing one’s mental health. Personal Informatics (PI) systems augmented by algorithms and sensors have become popular ways to help users collect and reflect on data about stress. While prediction algorithms in the PI systems are mainly for diagnostic purposes, few studies examine how the explainability of algorithmic prediction can support user-driven self-insight. To this end, we developed MindScope, an algorithm-assisted stress management system that determines user stress levels and explains how the stress level was computed based on the user’s everyday activities captured by a smartphone. In a 25-day field study conducted with 36 college students, the prediction and explanation supported self-reflection, a process to re-establish preconceptions about stress by identifying stress patterns and recalling past stress levels and patterns that led to coping planning. We discuss the implications of exploiting prediction algorithms that facilitate user-driven retrospection in PI systems.

One of the significant social problems emerging in modern society is mental illness, and a growing number of people are seeking psychological help. Art therapy is a technique that can alleviate psychological and emotional conflicts through creation. However, the expression of a drawing varies by individuals, and the subjective judgments made by art therapists raise the need to secure an objective assessment. In this paper, we present M2C (Multimodal classification with 2-stage Co-attention), a deep learning model that predicts stress from art therapy psychological test data. M2C employs a co-attention mechanism that combines two modalities-drawings and post-questionnaire answers-to complement the weaknesses of each, which corresponds to therapists' psychometric diagnostic processes. The results of the experiment show that M2C yielded higher performance than other state-of-the-art single- or multi-modal models, demonstrating the effectiveness of the co-attention approach that reflects the diagnosis process.

Spatio-temporal models, which combine GNNs (Graph Neural Networks) and RNNs (Recurrent Neural Networks), have shown state-of-the-art accuracy in traffic speed prediction. However, we find that they consider the spatial and temporal dependencies between speeds separately in the two (i.e., space and time) dimensions, thereby unable to exploit the joint-dependencies of speeds in space and time. In this paper, with the evidence via preliminary analysis, we point out the importance of considering individual dependencies between two speeds from all possible points in space and time for accurate traffic speed prediction. Then, we propose an Individual Spatio-Temporal graph (IST-graph) that represents the Individual Spatio-Temporal dependencies (IST-dependencies) very effectively and a Spatio-Temporal Graph ATtention network (ST-GAT), a novel model to predict the future traffic speeds based on the IST-graph and the attention mechanism. The results from our extensive evaluation with five real-world datasets demonstrate (1) the effectiveness of the IST-graph in modeling traffic speed data, (2) the superiority of ST-GAT over 5 state-of-the-art models (i.e., 2-33% gains) in prediction accuracy, and (3) the robustness of our ST-GAT even in abnormal traffic situations.

Many global automakers strive to develop technologies towards the next-generation of intelligent transportation systems (ITS). One of the primary goals of ITS is predicting future traffic speeds to optimize a driver's route, which can lead to not only alleviating traffic flow but also increasing user satisfaction with an ITS service. While prior studies have applied deep learning models to traffic speed prediction and improved model performance, existing models did not well capture abrupt speed changes. In this paper, we propose a novel model, named as adversarial prediction of traffic speed (APOTS), based on adversarial training, data augmentation, and hybrid deep learning modeling. Through the experiments with real traffic data provided by Hyundai Motor Company, we demonstrate that APOTS effectively learns dynamics of traffic speed changes and predicts traffic speed up to 40% higher in accuracy than existing prediction models.

Cybersickness is one of the problems that undermines user experience in virtual reality. While many studies are trying to find ways to alleviate cybersickness, only a few have considered cybersickness through multimodal perspectives. In this paper, we propose a multimodal, attention-based cybersickness prediction model. Our model was trained based on a total of 24,300 seconds of data from 27 participants and yielded the F1-score of 0.82. Our study results highlight the potential to model cybersickness from multimodal sensory information with a high level of performance and suggest that the model should be extended using additional, diverse samples.

Pervasive symptoms of people with autism spectrum disorder (ASD), such as a lack of social and communication skills, are major challenges to be embraced in the workplace. Although much research has proposed VR training programs, their effectiveness is somewhat unclear, since they provide limited, one-sided interactions through fixed scenarios or do not sufficiently reflect the characteristics of people with ASD (e.g., preference for predictable interfaces, sensory issues). In this paper, we present VISTA, a VR-based interactive social skill training system for people with ASD. We ran a user study with 10 people with ASD and 10 neurotypical people to evaluate user experience in VR training and to examine the characteristics of people with ASD based on their physical responses generated by sensor data. The results showed that ASD participants were highly engaged with VISTA and improved self-efficacy after experiencing VISTA. The two groups showed significant differences in sensor signals as the task complexity increased, which demonstrates the importance of considering task complexity in eliciting the characteristics of people with ASD in VR training. Our findings not only extend findings (e.g., low ROI ratio, EDA increase) in previous studies but also provide new insights (e.g., high utterance rate, large variation of pupil diameter), broadening our quantitative understanding of people with ASD.

As an emotion plays an important role in people’s everyday lives and is often mirrored in their social media use, extensive research has been conducted to characterize and model emotions from social media data. However, prior research has not sufficiently considered trends of social media use—the increasing use of images and the decreasing use of text—nor identified the features of images in social media that are likely to be different from those in non-social media. Our study aims to fill this gap by (1) considering the notion of visual hints that depict contextual information of images, (2) presenting their characteristics in positive or negative emotions, and (3) demonstrating their effectiveness in emotion prediction modeling through an in-depth analysis of their relationship with the text in the same posts. The results of our experiments showed that our visual hint-based model achieved 20% improvement in emotion prediction, compared with the baseline. In particular, the performance of our model was comparable with that of the text-based model, highlighting not only a strong relationship between visual hints of the image and emotion, but also the potential of using only images for emotion prediction which well reflects current and future trends of social media use.

A growing number of organizations are hiring autistic individuals as they start to recognize the value of a neurodiverse workforce. Despite this trend, the lack of support for autistic employees in workplaces complicates their employment. However, little is known about how people around autistic individuals can support them to create pleasant employment experiences. In this work, we develop the concept of the Workplace Playbook VR to investigate how virtual reality (VR) can help autistic people develop their work-related social communication skills in partnership with people in their support network. Using a video prototype to present the concept, we interviewed 28 participants, including 10 autistic people and 18 members of their support networks, which included family members and professionals. Our interviews revealed that the Workplace Playbook VR program can provide common ground for autistic people and members of their support network to participate in more empathetic communication regarding workplace challenges. Despite the benefits, we identified the potential misuse of social communication skills training features of the VR program to correct the personal characteristics of autistic individuals. Furthermore, to cultivate inclusive workplace environments, we found the needs of VR development not only for autistic people but also for neurotypical employees to promote their understanding of autism and empathy toward autistic employees. We suggest VR designs that promote a sense of agency and self-advocacy for autistic employees, and autism awareness and acceptance training for neurotypical employees.

Recent research on creativity support tools (CST) adopts artificial intelligence (AI) that leverages big data and computational capabilities to facilitate creative work. Our work aims to articulate the role of AI in supporting creativity with a case study of an AI-based CST tool in fashion design based on theoretical groundings. We developed AI models by externalizing three cognitive operations (extending, constraining, and blending) that are associated with divergent and convergent thinking. We present FashionQ, an AI-based CST that has three interactive visualization tools (StyleQ, TrendQ, and MergeQ). Through interviews and a user study with 20 fashion design professionals (10 participants for the interviews and 10 for the user study), we demonstrate the effectiveness of FashionQ on facilitating divergent and convergent thinking and identify opportunities and challenges of incorporating AI in the ideation process. Our findings highlight the role and use of AI in each cognitive operation based on professionals’ expertise and suggest future implications of AI-based CST development.

In this paper, we propose a framework for embedding-based community detection on signed networks. It first represents all the nodes of a signed network as vectors in low-dimensional embedding space and conducts a clustering algorithm (e.g., k-means) on vectors, thereby detecting a community structure in the network. When performing the embedding process, our framework learns only the edges belonging to balanced triangles whose edge signs follow the balance theory, significantly excluding noise edges in learning. To address the sparsity of balanced triangles in a signed network, our framework learns not only the edges in balanced real-triangles but those in balanced virtual-triangles that are produced by our generator. Finally, our framework employs adversarial learning to generate more-realistic balanced virtual-triangles with less noise edges. Through extensive experiments using seven real-world networks, we validate the effectiveness of (1) learning edges belonging to balanced real/virtual-triangles and (2) employing adversarial learning for signed network embedding. We show that our framework consistently and significantly outperforms the state-of-the-art community detection methods in all datasets.

Movement is one of the key elements in virtual reality (VR) and significantly influences user experience. In particular, walking-in-place is a method of supporting movement in a limited space, and many studies are being conducted on its effective support. However, most studies have focused on forward movement despite many situations in which backward movement is needed. In this paper, we present the development of a prediction model for forward/backward movement while considering a user's orientation and the verification of the model's effectiveness. We built a deep learning-based model by collecting sensor data on the movement of the user's head, waist, and feet. We developed three realistic VR scenarios that involve backward movement, set three conditions (controller-based, treadmill-based, and model-based) for movement, and evaluated user experience in each condition through a study of 36 participants. As a result, the model-based condition showed the highest sensory sensitivity, effectiveness, and satisfaction and similar cognitive burden compared with the other two conditions. The results of our study demonstrated that movement support through modeling is possible, suggesting its potential for use in many VR applications.

Fashion style analysis is of the utmost importance for fashion professionals. However, it has an issue of having different style classification criteria that rely heavily on professionals’ subjective experiences with no quantitative criteria. We present FANCY (Fashion Attributes detectioN for Clustering stYle), a human-centered, deep learning-based framework to support fashion professionals’ analytic tasks using a computational method integrated with their insights. We work closely with fashion professionals in the whole study process to reflect their domain knowledge and experience as much as possible. We redefine fashion attributes, demonstrate a strong association with fashion attributes and styles, and develop a deep learning model that detects attributes in a given fashion image and reflects fashion professionals’ insight. Based on attribute-annotated 302,772 runway fashion images, we developed 25 new fashion styles (FANCY dataset 1). We summarize quantitative standards of the fashion style groups and present fashion trends based on time, location, and brand.

Because of the increasingly negative impacts of the echo chamber effect, such as the dissemination of fake news and political polarization occurring in social networking services (SNSs), considerable efforts are being made to mitigate this effect. Prior HCI studies have presented the development of user interfaces to display information that reflects various standpoints, with the aim of nudging people to consume information in a more objective fashion. However, these efforts still lack the ability to highlight the characteristics, generation processes, and negative effects of echo chambers, so they may not be effective in helping people become sufficiently aware of the echo chamber effect and those who are already in an echo chamber. In this paper, we present ChamberBreaker (CB), which has been designed to help increase a player's awareness of and preemptively respond to an echo chamber effect based on psychological concepts: inoculation, heuristics for judging, and gamification. Through a user study with 882 participants (control group: 446, experimental group: 436), we demonstrated the feasibility of our game-based methodology to support the awareness of the echo chamber effect and the importance of maintaining diverse perspectives when consuming information. Our findings highlight the externalization of psychological standpoints in mitigating an echo chamber effect and suggest design implications for system development---the consideration of demographics, playing time, and the connection to fake news recognition---for digital literacy education. You can play CB at http://tiny.cc/chamberbreaker (The game only works with Chrome.)

Stress is caused by a variety of events in our daily lives. By anticipating stressful situations, we can prepare and better cope with stressors when they actually occur. However, many past-centric personal informatics (PI) tools focus on capturing events that already happened and analyzing the data. In this work, we examine how anticipation — a future-centric self-tracking practice — could be used to manage daily stress levels. To address this, we built MindForecaster, a calendar- mediated stress anticipation application that allows users to expect stressful events in advance, generates activities to mitigate stress, and evaluates actual stress levels compared to previously estimated stress levels. In a 30-day deployment with 47 users, the users who explicitly planned and executed coping interventions reported reduced stress more than those who only expected stressful events. We suggest design implications for stress management by incorporating the properties of anticipation into current PI models.

This study examines the impact of the 'diversity' of product recommendations on the 'preference' of a customer, using online/offline data from a leading fashion company. First, through interviews with fashion professionals, we categorized the characteristics of customers into four types - gift, coordinator, carry-over, and trendsetter. Then, using a hybrid filtering method, we increased the accuracy and diversity of recommended products. We derived 13 salient features that reflect customer behavior based on the Purchase Funnel model and built a classification model that predicts a customer's preference rates. Second, we conducted two large-scale user tests with 20,000 real customers to verify the effectiveness of our recommendation system. Study results empirically demonstrated the importance of diversity of recommended products. The more diverse the product recommendations were, the higher the purchase rate, the average purchase amount, and the cross purchase rate were observed. In addition, we tracked the customers? purchase for two months after the user tests and found that diverse product exposure positively influenced customer retention (e.g., repurchase rate, amount).

As social network services (SNS) are expanding from friend-based to interest-based, users form a new type of relationships, namely interest-based relationships, with friends and others through social activities (e.g., likes, comments). Although such relationships are highlighted in the common-identity theory and have important values in theoretical and practical aspects, little evidence exists in the literature pertaining to the explanation of social activities as a central component for social network analysis and an association with friendship. In this paper, we build like networks in Instagram and analyze them through the lens of two salient aspects - friendship and interest - that constitute social networks. Our study results (1) show ambiguous interpretations of the like activities between users who are friends, based on the comparative analysis between friend- and non-friend-based like networks, and (2) demonstrate strong signals of the hashtag characterizing the interest-based relationships in users and content. Our research substantiates and gives insights on the common-identity theory applied in online social networks through data-driven, empirical analysis.

While people primarily communicate with text in mobile chat applications, they are increasingly using visual elements such as images, emojis, and memes. Using such visual elements could help users communicate clearly and make chatting experience enjoyable. However, finding and inserting contextually appropriate images during the chat can be both tedious and distracting. We introduce MilliCat, a real-time image suggestion system that recommends images that match the chat content within a mobile chat application (i.e., autocomplete with images). MilliCat combines natural language processing (e.g., keyword extraction, dependency parsing) and mobile computing (e.g., resource and energy-efficiency) techniques to autonomously make image suggestions when users might want to use images. Through multiple user studies, we investigated the effectiveness of our design choices, the frequency and motivation of image usage by the participants, and the impact of MilliCat on mobile chat experiences. Our results indicate that MilliCat’s real-time image suggestion enables users to quickly and conveniently select and display images on mobile chat by significantly reducing the latency in the image selection process (3.19 × improvement) and consequently more frequent image usage (1.8 ×) than existing solutions. Our study participants reported that they used images more often with MilliCat as the images helped them convey information more effectively, emphasize their opinion, express emotions, and have fun chatting experience.

Motivated by a success of generative adversarial networks (GAN) in various domains including information retrieval, we propose a novel signed network embedding framework, ASiNE, which represents each node of a given signed network as a low-dimensional vector based on the adversarial learning. To do this, we first design a generator G+ and a discriminator D+ that consider positive edges, as well as a generator G - and a discriminator D- that consider negative edges: (1) G+/G- aim to generate the most indistinguishable fake positive/negative edges, respectsupively; (2) D+/D aim to discriminate between real positive/negative edges and fake positive/negative edges, respectively. Furthermore, under ASiNE, we propose two new strategies for effective signed network embedding: (1) an embedding space sharing strategy for learning both positive and negative edges; (2) a fake edge generation strategy based on the balance theory. Through extensive experiments using five real-life signed networks, we verify the effectiveness of each of the strategies employed in ASiNE. We also show that ASiNE consistently and significantly outperforms all the state-of-the-art signed network embedding methods in all datasets and with all metrics in terms of accuracy of sign prediction.

Subjective well-being (SWB) is a well-studied, widely used construct that refers to how people feel and think about their lives as one of many comprehensive perspectives on well-being. Much research has analyzed the role and utilization of technologies to improve one's SWB; however, especially when it comes to user modeling, multifaceted and variational aspects of SWB are less frequently considered. This paper presents an analysis on identifying factors for smartphone-based data on SWB and modeling SWB changes, based on a four-month user study with 78 college students. Our regression analysis highlights the significance of user attributes (e.g., personality, self-esteem) on SWB and salient factors derived from smartphone data (e.g., time spent on campus, ratio of standing/sitting stationary, expenses) that significantly account for SWB. Our classification analysis shows the potential for detecting SWB changes with reasonable performance, as well as for improving a model to be more tailored to individuals.

Geographical maps encoded with rainbow color scales are widely used by climate scientists. Despite a plethora of evidence from the visualization and vision sciences literature about the shortcomings of the rainbow color scale, they continue to be preferred over perceptually optimal alternatives. To study and analyze this mismatch between theory and practice, we present a web-based user study that compares the effect of color scales on performance accuracy for climate-modeling tasks. In this study, we used pairs of continuous geographical maps generated using climatological metrics for quantifying pairwise magnitude difference and spatial similarity. For each pair of maps, 39 scientist-observers judged: i) the magnitude of their difference, ii) their degree of spatial similarity, and iii) the region of greatest dissimilarity between them. Besides the rainbow color scale, two other continuous color scales were chosen such that all three of them covaried two dimensions (luminance monotonicity and hue banding), hypothesized to have an impact on task performance. We also analyzed subjective performance measures, such as user confidence, perceived accuracy, preference, and familiarity in using the different color scales. We found that monotonic luminance scales produced significantly more accurate judgments of magnitude difference but were not superior in spatial comparison tasks, and that hue banding had differential effects based on the task and conditions. Scientists expressed the highest preference and perceived confidence and accuracy with the rainbow, despite its poor performance on the magnitude comparison tasks. We also report on interesting interactions among stimulus conditions, tasks, and color scales, that lead to open research questions.

The purpose of this paper is to understand the factors involved when a human judges the credibility of information and to develop a classification model for weblogs, a primary source of information for many people. Considering both computational and human-centered approaches, we conducted a user study designed to consider two cognitive procedures--(1) visceral, behavioral and (2) reflective assessments--in the evaluation of information credibility. The results of the 80-participant study highlight that human cognitive processing varies according to an individual's purpose and that humans consider the structures and styles of content in their reflective assessments. We experimentally proved these findings through the development and analysis of classification models using 16,304 real blog posts written by 2,944 bloggers. Our models yield greater accuracy and efficiency than the models with well-known best features identified in prior research

As the number of TV channels increases, it is becoming important to recommend TV shows that users prefer to watch. To this end, we investigate the inherent characteristics of implicit feedback given in the TV show domain, and identify the challenges for building an effective TV show recommendation. Based on the unique characteristics, we define a user's watchable interval, the most important and novel concept in understanding users' true preferences. In order to reflect this new concept into the TV show recommendation, we propose a novel framework based on collaborative filtering. Our framework is composed of (1) preference estimation based on a user's watchable interval, (2) preference prediction based on confidence exploiting watchable episodes, and (3) top-N recommendation considering TV show's staying and remaining times. Using a real-world TV show dataset, we demonstrate that our framework effectively solves the challenges and significantly outperforms other existing state-of-the-art methods.

We present Reeboc that combines machine learning and k-means clustering to analyze the conversation of a chat, extract different emotions or topics of the conversation, and recommend emojis that represent various contexts to the user. Instead of simply analyzing a single input sentence, we consider recent sentences exchanged in a conversation. we performed a user study with 17 participants in 8 groups in a realistic mobile chat environment. Participants spent the least amount of time in identifying and selecting the emojis of their choice with Reeboc (38% faster than without emoji recommendation).

Although virtual reality (VR) programs to provide firefighting training continue to be developed and adopted, our investigation with 15 firefighters indicates that a current VR training system tends to convey behavioral tips and does not sufficiently reflect actual firefighters’ needs and realities in the field. It often provides somewhat simplified fire simulations and actually lowers the effectiveness of the training. In this paper, we employ Human-Computer Interaction (HCI) methods to examine and identify core elements in firefighting scenarios and develop a VR system that incorporates such elements. We evaluate our system with respect to presence and three design components of the VR simulation (i.e., reality, meaning, play) through a user study with 22 participants. Our study results demonstrate greater user experience and perception toward the four elements in firefighting training with our VR system compared to the existing one. We discuss design implications (e.g., move control, degree of freedom, sight hindrance by smoke, unexpected events) of our study that are expected to help implement and provide an effective VR training system for firefighters.

With the development of smart devices and the popularization of social media, communication between users is becoming increasingly active in the online space, revitalizing a new relationships formed and developed. Individuals are connecting with various people, which makes many companies actively employ social media to communicate with existing and potential customers for marketing. While social networks comprise a mix of various activities such as liking, commenting, tagging, and following, little is known about the way in which liking and tagging characterize one's social networks. In this paper, we define two types of a like network: tag-based like network, non-tag-based like network and present a comparative analysis with regard to their structural and temporal aspects. Our study results show (1) a significant difference in the network size and the degree of components between the two networks and (2) the potential of the tag-based like network to have more followers than the non-tag-based like network. We highlight how the tag-based like network can be utilized to find users with the same interests, supporting additional, interest based social connection and interaction. Our study insights are expected to be developed and utilized as a web service system.

Scientists often use specific data analysis and presentation methods familiar within their domain. But does high familiarity drive better analytical judgment? This question is especially relevant when familiar methods themselves can have shortcomings: many visualizations used conventionally for scientific data analysis and presentation do not follow established best practices. This necessitates new methods that might be unfamiliar yet prove to be more effective. But there is little empirical understanding of the relationships between scientists' subjective impressions about familiar and unfamiliar visualizations and objective measures of their visual analytic judgments. To address this gap and to study these factors, we focus on visualizations used for comparison of climate model performance. We report on a comprehensive survey-based user study with 47 climate scientists and present an analysis of: i) relationships among scientists' familiarity, their perceived levels of comfort, confidence, accuracy, and objective measures of accuracy, and ii) relationships among domain experience, visualization familiarity, and post-study preference.

The Massive Open Online Courses (MOOC) have experienced rapid development. However, high dropout rate has become a salient issue. Many studies have attempted to understand this phenomenon; other have explored mechanisms for enhancing retention. For instance, social media has been used to improve student engagement and retention. However there is a lack of (1) empirical studies of social media use and engagement compared to embedded MOOC forums; and (2) rationales for social media use from both instructors' and students' perspectives. We addressed these open issues through the collection and analysis of real usage data from three MOOC forums and their associated social media (i.e., Facebook) groups as well as conducting interviews of instructors and students. We found that students show higher engagement and retention in social media than in MOOC forums, and identified both instructors' and students' perspectives that lead to the results. We discuss design implications for future MOOC platforms.

Understanding the impacts and consequences of information and communications technology in local communities through theoretical and practical approaches has long been studied. Among different types of technologies, mobile technology suggests new opportunities for community informatics because a growing number of people across different age groups have adopted smartphones, which have become an indispensable part of people's daily lives. Because mobile technology transcends the limitations of time and place, it expands the ways of accessing and interacting with local community information and lowers the barrier to participation. In this paper, we present our ongoing initiatives in community informatics including digital cultural heritage and local volunteer efforts mediated by mobile technology. We highlight how mobile technology, together with increased mobility, immediacy and social presence, shows a significant influence on local communities with respect to community identity, awareness and participation and also social support networks.

The emergence of social media has had a significant impact on how people communicate and socialize. Teens use social media to make and maintain social connections with friends and build their reputation. However, the way of analyzing the characteristics of teens in social media has mostly relied on ethnographic accounts or quantitative analyses with small datasets. This paper shows the possibility of detecting age information in user profiles by using a combination of textual and facial recognition methods and presents a comparative study of 27K teens and adults in Instagram. Our analysis highlights that (1) teens tend to post fewer photos but highly engage in adding more tags to their own photos and receiving more Likes and comments about their photos from others, and (2) to post more selfies and express themselves more than adults, showing a higher sense of self-representation. We demonstrate the application of our novel method that shows clear trends of age differences as well as substantiates previous insights in social media.

Timebanking is a service-based community currency, built on the principle that everyone's time is valued equally. It has potential for community building and reenergizing neighborhoods, but it faces several adoption challenges. We report on the largest investigation of timebanking practices to date by analyzing a combination of service exchange records from the three largest hOurworld timebanks with over 3,500 members with 33,000 completed service exchanges, and a survey of 446 members of over 120 hOurworld timebanks. Our findings suggest that the ideal of 'equal time, equal value' that is at the foundation of timebanking is a source of tension between members with instrumental versus idealistic and altruistic motivations. We suggest that future peer-to-peer systems must incorporate different rewards and incentives in order to accommodate users with different motivations.

Commercial peer-to-peer service exchange businesses, such as AirBnB, Lyft and TaskRabbit, are expanding rapidly, but their non-profit counterparts are lagging behind. We conducted a field study of the most prominent of these, timebanking; a system in which 'time dollars' are earned and spent by people providing services for and receiving them from each other. Our study exposed problems with the very metaphor of banking itself, which deter participation. In this paper we discuss how these problems can be tackled with user experience design for systems supporting timebanking. Our design ideas emphasize the personal and social benefits of participation, and avoid such unappealing concepts as debt and neediness that the timebanking metaphor falls afoul of.

One goal of local communities is to create and reinforce community identity by connecting residents to their local heritage. Technologies have enabled and facilitated the creation and consumption of digitized history content provided by official history institutions as well as individuals. Although much research has been conducted to understand technical and social aspects of digital cultural heritage, little empirical research has investigated how people perceive, experience, and interact with community content that is socially generated and tied to locations, particularly with respect to building community heritage. To address this, we developed a mobile application called Lost State College (LSC) and conducted a user study with 34 local residents. The study results indicate that meaningful historic places evoked special attention from the participants, and that those who have lived in the community longer tended to contribute more to the community heritage effort. Participants utilized social features as a way of learning local history, reflecting personal experiences and stories, and co-creating rich layers of local history information from their perspectives.

Social media platforms such as Twitter and Facebook are commonly used to disseminate up-to-date news information, but they also contain a lot of noise and irrelevant content. The contents of social media platforms are typically filtered by followship or friendship oriented relationships, and is almost always driven by trending news topics at the national scale, making it difficult for users to gather useful information that is most pertinent to a local community context. Research has utilized content analysis techniques to gain insights on the sentiment expressed about political topics on social media sites. However, there has been little attempt to understand how users would perceive this information if opinions and sentiments about news topics were externalized and made aware to them. We designed Community Poll, a smartphone application that aggregates local news feeds with relevant tweets about the local news topics. A Public Attitude Meter is calculated based on the sentiment score of the tweets for each of the local news topic presented in the system. We conducted a 2-week deployment with 16 users about their perception of the system. The users reported that Community Poll helps them digest locally relevant news topics, and quickly gather public opinions associated with the topics. They reported that being aware of public sentiment encouraged them to more actively participate in discussions on social media. Curiosity about a score-based representation is an important element that drove them to consume local news topics that they wouldn’t otherwise be exposed to. Interestingly, although being aware of public sentiment served to reaffirm people’s positions on the local topics, users expressed concerns about how sentiment awareness might bias other people’s judgments regarding news topics.

Being aware of local community information is critical to maintaining civic engagement and participation. The use of online news and microblog content to create and disseminate community information has long been studied. However, interactions in the online spaces dedicated to local communities tend to only garner very limited usage, and people often do not consider microblog content as a meaningful source of local community information. Local News Chatter (LNC) was designed to address these challenges by augmenting local news feeds with microblog content and presenting them in a tag cloud that displays news topics of varying popularity with different tag sizes. Our study with 30 local residents highlights that LNC increases the visibility of hyperlocal community news information and successfully utilizes microblog as an additional information layer. LNC also increases one’s community awareness and shows the potential for leveraging community knowledge as a deliberation platform for local topics.

Just as coordinate-oriented location-based applications have exploded recently with mapping services, new semantic location services will be critical for the next wave of killer applications. People are going to want everyday applications to have location-awareness that goes beyond simple numerical latitude and longitude. Loci is a new semantic location service layer that employs user feedback to bridge the gap between machine-learned and human-defined places. Advances in place learning techniques have provided us the tools to detect nearly 95% of the visits we make to places and the distances we travel. The difficulty of recovering the remaining 5% comes from designing parameters that work for every user in every place. Based on a user study with 29 participants over three weeks, we show that the level of user feedback required by the service is acceptable and most of the users are willing to provide help to improve their experiences with the service. Our results suggest that user feedback has the potential to significantly improve semantic location services, but requires well-timed prompting mechanisms to improve the quality of the feedback.