Human Computer Interaction
Human-computer interaction (HCI) is a multidisciplinary field of study that focuses on how people interact with technology. This technology can range from physical artifacts such as sensors and wearable devices, to networked systems such as social media and online communities and algorithms (such as AI and generative AI). HCI can be applied to almost all aspects of society, in domains such as computer-supported collaborative work, health, education, security/privacy, music, cyberpsychology, social computing and entertainment using qualitative, quantitative and data visualization methods.. The field also includes the study of technology use with special populations to examine accessibility and inclusive technology. This work is frequently published in top-tier venues such as ACM CHI, ACM CSCW, ACM ASSETS, TVCG and IEEE VIS.
|
Julie Ancis |
Research Areas: Cyberpsychology, human-computer interaction, tailored social media messaging, misinformation, online hate, social media infleuncers, disinformation, bias Cyberpsychological Impacts of Online Information The informational landscape and how individuals and groups engage with this information has implications for a broad range of human behaviors. Using qualitative and quantitative approaches grounded in psychology and computing, we analyze how both the text and visual components of social media messages are constructed to convey information about issues ranging from health crises to critical societal events. Our research investigates the impact of these messages on risk perception, decision making, proactive health behaviors, civil discourse and bias. This work has four major foci: 1) The evaluation of tailored and targeted online messaging. 2) Development of culturally informed online information and messages to enable adaptive behavior. 3) Testing the psychological and behavioral impact of culturally relevant messaging on perceptions and intentions. |
|
Mark Cartwright |
Research Areas: Machine listening, interactive machine learning, human-computer interaction, audio processing, music information retrieval, sound accessibility Tools for Accessible Sound Understanding and Creation Sound is all around us and in the media we consume. However, for the roughly 1 in 5 people who are to some degree deaf or hard of hearing (DHH), the information contained in this medium is not always accessible. In this project, we aim both to develop tools to increase the accessibility of information contained in sound and to develop tools to make the creation of audio artifacts more accessible. |
|
Aritra Dasgupta |
Research Areas: human-AI teaming, interactive visualization, visual analytics, responsible AI, humanmachine communication Human-AI Collaboration using Visual Analytics Our research group aims to address the question: “How can we bridge the reasoning gap between people & data-driven models?” As newer technologies such as large language models and generative AI continue to be developed and adopted in socio-technical applications, it will be increasingly critical to address this reasoning gap so that human stakeholders have appropriate levels of agency in interpreting, verifying and partaking in the model-backed claims and decision-making processes. Complex algorithms and AI models often drive decisions based on available data. Be it climate scientists simulating global warming, epidemiologists forecasting regional pandemic trajectories, or digital interfaces recommending products to consumers, data needs to be modeled reliably to provide real analytical value. However, developing and interpreting data-driven models, like those resulting from machine learning or scientific simulations, necessitate substantial human effort to understand the context and significance of patterns in the data. To address these challenges, we conceptualize, design and evaluate visualization as an evidence-based communication medium between humans and complex AI models. The techniques we build integrate automated methods such as statistics and machine learning with optimal visualizations to enable human judgment and reasoning about AI-detected patterns. This integration ensures that we leverage the best of both worlds: computational power for fast extraction of patterns (e.g., a predictive model trained on millions of transactions for flagging suspicious individuals) and our perceptual and cognitive faculties for letting experts transparently reason about the context and significance of the patterns (e.g., financial analysts visualizing past and present behavior of individuals to understand the context of predictions). The artifacts of our research are evaluated using a mix of quantitative (e.g., accuracy metric) and qualitative (e.g., design study and case study) methods. My research methodology entails participatory design of visualizations and interactive systems, empirical evaluation of techniques through user studies and observational techniques to study how people interact with data and AI models. |
|
Amy Hoover |
Research Areas: Optimization Methods in Quality Diversity While many optimization techniques in evolutionary computation maximize fitness with respect to one or more objectives, often such searches restrict the set of candidate solutions to those with objective values lying along the Pareto front of optimality. Instead, algorithms in Quality Diversity (QD) fully explore these objectives by specifying them explicitly as dimensions (called behavioral characteristics) that are characterized by their genomic, phenotypic or behavioral traits. Such exploration can not only generate a large collection of high-performing solutions, but with well-chosen dimensions can potentially find higher performing solutions than pure-objective based searches alone. Predictive Analytics on a Cluster of Computers Predicting viral events in social networks in real time is challenging as events can unfold in a matter of months, weeks, days or even hours. We have been developing a system of hardware and software for real-time analytics, specifically targeting the prediction of viral events in social networks using a cluster of computers. One of the key enabling technologies is inertial spectral graph partitioning, developed in collaboration with NASA and Berkeley Lab. We have successfully implemented the framework to predict viral events in social networks, achieving over 70% accuracy on large-scale dynamic temporal Wikipedia graphs. We are currently working on further improving prediction accuracy. |
|
Sooyeon Lee |
Research Areas: Human-computer interaction, accessibility, human-AI interaction Design and Evaluation of Accessible AI Technologies for Users with Disabilities Over one billion people in the world live with some type of disability. Many of them experience barriers in accessing information or using technologies, which can limit social interactions in both physical and digital spaces. Our work investigates the diversity of users, explores and leverages emerging technologies and adopts human-centered AI and inclusive design approaches in the design and evaluation of new AI based systems and applications that address accessibility barriers. Our research focuses on investigating and designing non-visual interaction for the community of blind users and non-audio and nonspeech interaction for the community of deaf and hard of hearing users. |
|
Alisha Pradhan |
Research Areas: Human-computer interaction, human-AI interaction, inclusive design, aging, accessibility, underrepresented populations Towards Inclusive Design of Emerging Technologies In today’s world, our dependence on technology has become so profound that it influences nearly every aspect of our daily lives. But when it comes to design of such technologies, who do product designers typically consider as users? More importantly, who do they exclude as users (even though, inadvertently)? And what barriers and challenges do those users, who are rarely considered in the technology design process, experience when using these systems? What harmful assumptions or stereotypes do our technologies encode about them? These are some of the questions that guide our research on inclusive design of emerging technologies. Older adults represent one group, typically underrepresented in technology design, where their perspectives and preferences are rarely included. By using user-centered design approaches, we closely work with older individuals in design and development of technologies. We examine the benefits and barriers posed by existing technologies, as well as adopt participatory approaches to design technologies for and with older individuals. |
|
Erin J.K. Truesdell |
Research Areas: Human-computer interaction, games, game design, serious games, extended reality, tangible interfaces, collaboration, play Understanding Collaboration with Tangible Interfaces Our research centers on how to design novel tangible interfaces to support multi-user collaboration with a particular focus on playful and creative applications. We are interested in understanding how people work together using a variety of tangible and wearable interfaces to achieve goals, produce creative work, develop interpersonal skills and relationships and learn new information through interaction. Insights gained from this research have applications in entertainment and beyond, in fields such as computer-supported collaborative work, formal and informal learning and human-AI interaction. |
|
Yvette Wohn |
Research Areas: Human-computer interaction, social media, digital economy, online self-presentation Social Media and Wellbeing This research investigates the relationship between psychological well-being and technology usage, especially how social media can play a role in facilitating both hate and social support. This includes the study of individualistic behavior—such as understanding how people form and present their identity online—as well as collective behavior, such as how people engage in collective action and activism. New Digital Economies The metaverse may seem like a new phenomenon but research on virtual spaces has been going on for decades. In this research, we focus on systems with novel digital economies in virtual environments that have unique digital currency, such as online games. Our research examines spending behaviors and exchange patterns of virtual goods in games and other alternative financial platforms and how these activities are tied with creative content generation. |
