Georgia Tech @ CHI 2017

CHI is the Association for Computing Machinery’s annual Conference on Human Factors in Computing Systems



Georgia Tech Among Top Participating Institutions at CHI 2017

This week, the Mile High City is transforming into the “Mile CHI City” as scores of researchers from across the country and around the globe descend on Denver this week (May 8 – 11) for the Association for Computing Machinery’s annual Conference on Human Factors in Computing Systems (CHI).

Among the throng is the 29-person Georgia Tech contingent. Better known as CHI, the event is the premiere international ACM conference on human-computer interaction (HCI) attracting researchers from around the globe.

The faculty and students representing Georgia Tech at CHI 2017 are from the College of Computing, Ivan Allen College of Liberal Arts, and the GVU Center. Together, they have 17 research papers and notes accepted as part of the technical program.

Georgia Tech also boasts an expansive alumni network at CHI. These researchers, along with the faculty and student contingent, will take part in a variety of workshops, panels, courses, and other parts of the CHI technical program.

Overall, Georgia Tech ranks in the top ten for participation among the 500 or so institutions represented at CHI 2017.


Papers and Notes


Women’s Safety in Public Spaces: Examining the Efficacy of Panic Buttons in New Delhi

We present a qualitative inquiry through the lens of feminist Human-Computer Interaction (HCI) into women’s perceptions of personal safety in New Delhi, India. Since a brutal gang-rape incident took place in Delhi in December 2012 and received global attention, women’s safety has been the focus of much attention India-wide. In April 2016, the Indian government issued a mandate that all mobile phones sold in India 2017 onwards must include a panic button for women’s safety. We draw on interview and survey data to examine women’s responses to the mandate, also investigating what factors influence their perceptions of safety, positively and negatively. Our findings indicate that women’s sense of safety may be deconstructed into a multitude of factors–personal, public, social, technological–that must align for this sense of safety to be preserved. We then discuss the implications these factors have for the success and (re-)design of the panic button and similar interventions.

Topics: Safety; Gender; India; HCI4D; Feminist HCI

Naveena Karusala (Georgia Institute of Technology)
Neha Kumar (Georgia Institute of Technology)


Rice Today, Roti Tomorrow: Diets and Diabetes in Urban Indian Households

In India, where diabetes is a growing concern and approximately 69 million are affected, we investigate the factors that influence diet management, a critical component of living with the disease. Taking the middle-income diabetes-affected household as our unit of analysis, we use a combination of semi-structured interviews and a design probe to understand if and how diets are monitored, tailored, and balanced. We research the various information-seeking behaviors of our participants and their culturally situated approaches to food and eating. Our findings illuminate how contextual nuances shape individuals’ beliefs around dealing with diabetes and the ways in which family, friends, and broader social networks influence dietary decisions. We conclude by offering a framework of Learning-Being-Doing to inform the holistic design of technologies for managing diets and diabetes.

Topics: Diabetes; Diets; India; Qualitative Methods

Jasmine Hentschel (THRIVE)
Samyukta Manjayya Sherugar (Georgia Institute of Technology)
Rui Zhou (Georgia Institute of Technology)
Vaishnav Kameswaran (University of Michigan)
Rajesh Chandwani (Indian Institute of Management)
Neha Kumar (Georgia Institute of Technology)


Situated Anonymity: Impacts of Anonymity, Ephemerality, and Hyper-Locality on Social Media

Anonymity, ephemerality, and hyper-locality are an uncommon set of features in the design of online communities. However, these features were key to Yik Yak’s initial success and popularity. In an interview-based study, we found that these three features deeply affected the identity of the community as a whole, the patterns of use, and the ways users committed to this community. We conducted interviews with 18 Yik Yak users on an urban American university campus and found that these three focal design features contributed to casual commitment, transitory use, and emergent community identity. We describe situated anonymity, which is the result of anonymity, ephemerality, and hyper-locality coexisting as focal design features of an online community. This work extends our understanding of use and identity-versus-bond based commitment, which has implications for the design and study of other atypical online communities.

Topics: online communities; anonymity; ephemerality; hyper-locality; commitment; transitory use; community identity

Ari Schlesinger (Georgia Institute of Technology)
Eshwar Chandrasekharan (Georgia Institute of Technology)
Christina A Masden (Georgia Institute of Technology)
Amy S Bruckman (Georgia Institute of Technology)
W Keith Edwards (Georgia Institute of Technology)
Rebecca E Grinter (Georgia Institute of Technology)


Understanding the Cost of Driving Trips

Driving is the second highest expense for the average American household. Yet few people know the total cost of owning and operating their vehicles, and most cannot estimate accurately how much a common driving trip (like a daily commute) costs. There are an increasing number of viable alternatives for personal transportation, such as car services (e.g. Uber, Lyft), in addition to ridesharing, transit, biking, and walking. Cost is one factor in transportation mode choice, and awareness of the cost of driving is useful in making better informed decisions. To bridge this awareness gap, we built and deployed a system that makes the total cost of each driving trip (including depreciation, maintenance, insurance, and fuel) visible to the user. After this intervention, participants were able to more accurately and confidently estimate costs of their driving commutes, and transfer this knowledge to other trips for which they had not seen a cost.

Topics: Transportation; driving; mode choice; personal informatics

Caleb Southern (Georgia Institute of Technology)
Yunnuo Cheng (Georgia Institute of Technology)
Cheng Zhang (Georgia Institute of Technology)
Gregory D Abowd (Georgia Institute of Technology)


Locating the Internet in the Parks of Havana

Since March 2015, the public squares of Havana have been transformed from places where people stroll and children play to places where crowds gather to try to connect to the internet at all hours of the day and night. We present a field investigation of public WiFi hotspots in Havana, Cuba, and examine the possibilities of internet access these limited and expensive hotspots present to individuals, many of who are experiencing the internet for the first time. Drawing on fieldwork conducted in 2015-2016, we underscore the reconfigurations that have resulted from this access, as evolving internet users reconfigure their interactions with place, time, and individuals in their efforts to locate the internet. We also discuss the implications our findings have for the design of internet access interventions in Cuba and in other low-resource environments across the world, as well as the broader implications for social computing across diverse geographies.

Topics: Cuba; internet; WiFi; place; HCI4D; social computing

Michaelanne Dye (Georgia Institute of Technology)
David Nemer (University of Kentucky)
Laura Pina (University of Washington)
Nithya Sambasivan (Google)
Amy S Bruckman (Georgia Institute of Technology)
Neha Kumar (Georgia Institute of Technology)


Self-tracking for Mental Wellness: Understanding Expert Perspectives and Student Experiences

Previous research suggests an important role for self-tracking in promoting mental wellness. Recent studies with college student populations have examined the feasibility of collecting everyday mood, activity, and social data. However, these studies do not account for students’ experiences and challenges adopting self-tracking technologies to support mental wellness goals. We present two studies conducted to better understand self-tracking for stress management and mental wellness in student populations. First, focus groups and card sorting activities with 14 student health professionals reveal expert perspectives on the usefulness of tracking for three scenarios. Second, an online survey of 297 students examines personal experiences with self-tracking and attitudes toward sharing self-tracked data with others. We draw on findings from these studies to characterize students’ motivations, challenges and preferences in collecting and viewing self-tracked data related to mental wellness, and we compare findings between students with diagnosed mental illnesses and those without. We conclude with a discussion of challenges and opportunities in leveraging self-tracking for mental wellness, highlighting several design considerations.

Topics: Self-monitoring; self-tracking; personal informatics; Quantified Self; mental health; mental wellness; health communication; patient–clinician communication

Christina Kelley (Georgia Institute of Technology)
Bongshin Lee (Microsoft Research)
Lauren Wilcox (Georgia Institute of Technology)


Supporting Families in Reviewing and Communicating about Radiology Imaging Studies

Diagnostic radiology reports are increasingly being made available to patients and their family members. However, these reports are not typically comprehensible to lay recipients, impeding effective communication about report findings. In this paper, we present three studies informing the design of a prototype to foster patient–clinician communication about radiology report content. First, analysis of questions posted in online health forums helped us identify patients’ information needs. Findings from an elicitation study with seven radiologists provided necessary domain knowledge to guide prototype design. Finally, a clinical field study with 14 pediatric patients, their parents and clinicians, revealed positive responses of each stakeholder when using the prototype to interact with and discuss the patient’s current CT or MRI report and allowed us to distill three use cases: co-located communication, preparing for the consultation, and reviewing radiology data. We draw on our findings to discuss design considerations for supporting each of these use cases.

Topics: Families; Adolescents; Radiology Report; Patient–Doctor Communication

Matthew K Hong (Georgia Institute of Technology)
Clayton Feustel (Georgia Institute of Technology)
Meeshu Agnihotri (Georgia Institute of Technology)
Max Silverman (Georgia Institute of Technology)
Stephen F Simoneaux (Children’s Healthcare of Atlanta)
Lauren Wilcox (Georgia Institute of Technology)


Intersectional HCI: Engaging Identity through Gender, Race, and Class

Understanding users becomes increasingly complicated when we grapple with various overlapping attributes of an individual’s identity. In this paper we introduce intersectionality as a framework for engaging with the complexity of users’—and authors’—identities, and situating these identities in relation to their contextual surroundings. We conducted a meta-review of identity representation in the CHI proceedings, collecting a corpus of 140 manuscripts on gender, ethnicity, race, class, and sexuality published between 1982-2016. Drawing on this corpus, we analyze how identity is constructed and represented in CHI research to examine intersectionality in a human-computer interaction (HCI) context. We find that previous identity-focused research tends to analyze one facet of identity at a time. Further, research on ethnicity and race lags behind research on gender and socio-economic class. We conclude this paper with recommendations for incorporating intersectionality in HCI research broadly, encouraging clear reporting of context and demographic information, inclusion of author disclosures, and deeper engagement with identity complexities.

Topics: Intersectional HCI; gender; ethnicity; race; class, socio-economic status; intersectionality; identity

Ari Schlesinger (Georgia Institute of Technology)
W Keith Edwards (Georgia Institute of Technology)
Rebecca E Grinter (Georgia Institute of Technology) Designing a Workflow-Based Scheduling Agent with Humans in the Loop

Although we may complain about meetings, they are an essential \ part of an information worker’s work life. Consequently, \ busy people spend a significant amount of time scheduling \ meetings. We present, a system that provides \ fast, efficient scheduling through structured workflows. Users \ interact with the system via email, delegating their scheduling \ needs to the system as if it were a human personal assistant. \ Common scheduling scenarios are broken down using well-defined \ workflows and completed as a series of microtasks \ that are automated when possible and executed by a human \ otherwise. Unusual scenarios fall back to a trained human \ assistant executing an unstructured macrotask. We describe \ the iterative approach we used to develop, and \ share the lessons learned from scheduling thousands of meetings \ during a year of real-world deployments. Our findings \ provide insight into how complex information tasks can be \ broken down into repeatable components that can be executed \ efficiently to improve productivity.

Topics: Scheduling; microtask; macrotask; crowdsourcing; conversational agent; assistant

Justin B Cranshaw (Microsoft Research)
Emad Elwany (Microsoft Research)
Todd Newman (Microsoft Reserach)
Rafal Kocielnik (University of Washington)
Bowen Yu (University of Minnesota)
Sandeep Soni (Georgia Institute of Technology)
Jaime Teevan (Microsoft Research)
Andres Monroy-Hernandez (Microsoft Research)


Goodbye Text, Hello Emoji: Mobile Communication on WeChat in China

We present a qualitative study of mobile communication via WeChat in Southern China, focusing on the rapid proliferation of emoji and stickers and the lessening dependence on text. We use interview and observation data from 30 participants to investigate how rural, small town, and urban Chinese adults creatively and innovatively balance the use of emoji, stickers, and text in their mobile communication practices. We also discuss design implications of our research for the field of HCI, offering ways of leveraging the non-textual communication practices that we uncover, in scenarios where purely text-based communication may not suffice.

Topics: China; WeChat; Mobile; Emoji; Stickers; Qualitative Methods

Rui Zhou (Georgia Institute of Technology)
Jasmine Hentschel (THRIVE)
Neha Kumar (Georgia Institute of Technology)


A Social Media Based Index of Mental Well-Being in College Campuses

Psychological distress in the form of depression, anxiety and other mental health challenges among college students is a growing health concern. Dearth of accurate, continuous, and multi-campus data on mental well-being presents significant challenges to intervention and mitigation efforts in college campuses. We examine the potential of social media as a new “barometer” for quantifying the mental well-being of college populations. Utilizing student-contributed data in Reddit communities of over 100 universities, we first build and evaluate a transfer learning based classification approach that can detect mental health expressions with 97% accuracy. Thereafter, we propose a robust campus-specific Mental Well-being Index: MWI. We find that MWI is able to reveal meaningful temporal patterns of mental well-being in campuses, and to assess how their expressions relate to university attributes like size, academic prestige, and student demographics. We discuss the implications of our work for improving counselor efforts, and in the design of tools that can enable better assessment of the mental health climate of college campuses.

Topics: college mental health; Reddit; social media; transfer learning

Shrey Bagroy (Indian Institute of Technology)
Ponnurangam Kumaraguru (Indian Institute of Technology)
Munmun De Choudhury (Georgia Institute of Technology)


Modeling and Understanding Visual Attributes of Mental Health Disclosures in Social Media

Content shared on social media platforms has been identified to be valuable in gaining insights into people’s mental health experiences. Although there has been widespread adoption of photo-sharing platforms such as Instagram in recent years, the role of visual imagery as a mechanism of self-disclosure is less understood. We study the nature of visual attributes manifested in images relating to mental health disclosures on Instagram. Employing computer vision techniques on a corpus of thousands of posts, we extract and examine three visual attributes: visual features (e.g., color), themes, and emotions in images. Our findings indicate the use of imagery for unique self-disclosure needs, quantitatively and qualitatively distinct from those shared via the textual modality: expressions of emotional distress, calls for help, and explicit display of vulnerability. We discuss the relationship of our findings to literature in visual sociology, in mental health self disclosure, and implications for the design of health interventions.

Topics: social media; mental health; Instagram; visual attributes

Lydia Manikonda (Arizona State University)
Munmun De Choudhury (Georgia Institute of Technology)


Sidestepping the Elephant in the Classroom: Using Culturally Localized Technology To Teach Around Taboos

Cultural taboos can restrict student learning on topics of critical importance. In India, such taboos have led multiple states to ban materials intended to educate youth about HIV, putting millions at risk. We present the design of TeachAIDS, a software application that leverages cultural insights, learning science, and affordances of technology to provide comprehensive HIV education while circumventing taboos. Using a mixed-methods evaluation, we demonstrate that this software leaves students with significantly increased knowledge about HIV and reduced stigma toward individuals infected with the virus. Validating the effectiveness of TeachAIDS in circumventing taboos, students report comfort in learning from the software, and it has since been deployed in tens of thousands of schools throughout India. The methodology presented here has broader implications for the design and implementation of interactive technologies for providing education on sensitive topics in health and other areas.

Topics: Taboo Topics; Education; HIV; AIDS; India; HCI4D

Piya Sorcar (Stanford University)
Benjamin Strauber (Stanford University)
Prashant Loyalka (Stanford University)
Neha Kumar (Georgia Institute of Technology)
Shelley Goldman (Stanford University)


Creating a Sociotechnical API: Designing City-Scale Community Engagement

Community engagement is to cities what user experience is to computing: it signifies a large category that simultaneously speaks to general qualities of interaction and to specific ways of doing that interaction. Recently, digital civics has emerged as a research area with a comprehensive approach to designing for civic encounters where community engagement is a primary concern for designing systems and processes that support broad civic interaction. In short, over the past year, we worked with municipal officials, service providers, and city residents to design a community engagement playbook detailing best practices for city-scale engagement. The playbook, as well as the collaborative process that produced it, provides a roadmap for thinking through the kinds of systems that might populate the design space of city-scale digital civics. This paper details our design-led research process and builds on emerging literature on designing for digital civic interaction.

Topics: Digital Civics; Participatory Design; Publics; Community Engagement

Mariam Asad (Georgia Institute of Technology)
Christopher A Le Dantec (Georgia Institute of Technology)
Becky Nielsen (IT University of Copenhagen)
Kate Diedrick (Solidarity Research Cooperative)


Multimodal Classification of Moderated Online Pro-Eating Disorder Content

Social media sites are challenged by both the scale and variety of deviant behavior online. While algorithms can detect spam and obscenity, behaviors that break community guidelines on some sites are difficult because they have multimodal subtleties (images and/or text). Identifying these posts is often regulated to a few moderators. In this paper, we develop a deep learning classifier that jointly models textual and visual characteristics of pro-eating disorder content that violates community guidelines. Using a million Tumblr photo posts, our classifier discovers deviant content efficiently while also maintaining high recall (85%). Our approach uses human sensitivity throughout to guide the creation, curation, and understanding of this approach to challenging, deviant content. We discuss how automation might impact community moderation, and the ethical and social obligations of this area.

Topics: content moderation; social media; pro-eating disorder; deep learning; computer vision; deviant behavior; Tumblr

Stevie Chancellor (Georgia Institute of Technology)
Yannis Kalantidis (Yahoo! Inc)
Jessica Annette Pater (Georgia Institute of Technology)
Munmun De Choudhury (Georgia Institute of Technology)
David A. Shamma (Centrum Wiskunde & Informatica)


The Bag of Communities: Identifying Abusive Behavior Online with Preexisting Internet Data

Since its earliest days, harassment and abuse have plagued the Internet. Recent research has focused on in-domain methods to detect abusive content and faces several challenges, most notably the need to obtain large training corpora. In this paper, we introduce a novel computational approach to address this problem called Bag of Communities (BoC)—a technique that leverages large-scale, preexisting data from other Internet communities. We then apply BoC toward identifying abusive behavior within a major Internet community. \ Specifically, we compute a post’s similarity to 9 other communities from 4chan, Reddit, Voat and MetaFilter. We show that a BoC model can be used on communities “off the shelf” with roughly 75% accuracy—no training examples are needed from the target community.  \ A dynamic BoC model achieves 91.18% accuracy after seeing 100,000 human-moderated posts, and uniformly outperforms in-domain methods. Using this conceptual and empirical work, we argue that the BoC approach may allow communities to deal with a range of common problems, like abusive behavior, faster and with fewer engineering resources.

Topics: social computing; online communities; abusive behavior; moderation; machine learning

Eshwar Chandrasekharan (Georgia Institute of Technology)
Mattia Samory (University of Padua)
Anirudh Srinivasan (Georgia Institute of Technology)
Eric Gilbert (Georgia Institute of Technology)


Tap the “Make This Public” Button: A Design-Based Inquiry into Issue Advocacy and Digital Civics

This paper examines the strategies of cycling advocates when deploying digital tools in their advocacy work as they support and create better cycling infrastructure and policies. Over the course of two years, we interviewed and conducted design-based fieldwork in two large U.S. cities with individuals and advocacy organizations, learning about the goals, motivations, and constraints that inform their work in their respective urban homes. Our design-based investigation and fieldwork advance a deeper, situated understanding of the role that computing technology plays when engaging across multiple sites of advocacy work. From this, we add detail to the connections across resources, identities, and issues and continue to advance the emerging area of digital civics, which seeks to design tools that support relational civic interactions across multiple categories of civic actors.

Topics: Digital advocacy; community computing; participatory design; civic engagement; publics; digital civics; cycling

Mariam Asad (Georgia Institute of Technology)
Christopher A Le Dantec (Georgia Institute of Technology)