Michigan scholars show strength at CHI

Three honorable mention papers, two teams of finalists in the Student Design Competition, and one faculty member being inducted into the CHI Academy are among the highlights of the U-M School of Information’s participation in the Association of Computing Machinery (ACM) Special Interest Group on Human-Computer Interaction (SIGCHI) conference taking place in Denver May 6-11, 2017. The ACM CHI Conference on Human Factors in Computing Systems is the premier international conference of Human-Computer Interaction.

Professor Paul Resnick will receive a SIGCHI CHI Academy Award at a ceremony on Sunday, May 7. Resnick is the Michael D. Cohen Collegiate Professor of Information and Associate Dean for Research and Faculty Affairs at UMSI. He was a pioneer in the field of recommender systems (sometimes called collaborative filtering). The GroupLens system he helped develop was awarded the 2010 ACM Software Systems Award. An article written with Eric Friedman, "The Social Cost of Cheap Pseudonyms" received the inaugural ACM EC Test of Time Award. His 2012 MIT Press book, co-authored with Robert Kraut, was titled Building Successful Online Communities: Evidence-based Social Design.

Two student teams from UMSI are among the finalists in the Student Design Competition, facing off on Wednesday, May 10. The projects and their team members are “AllergyBot: A Chatbot Technology Intervention for Young Adults with Food Allergies Dining Out” (Kehan Liao, Jingshu Zhao, Tianyi Liu, Chen Wang, Paris Hsu) and “Elevate: Ensuring Access to Food for Homeless Populations” (Abhraneel Sarma, Nisha Mohan).

Among the several papers being presented by UMSI faculty and doctoral students, three have been awarded Honorable Mentions.

Managing Uncertainty in Time Expressions for Virtual Assistants,” authored by UMSI doctoral student Xin Rong with collaborators Adam Fourney, Robin N Brewer (UMSI Presidential Post-doc, Fall '17), Meredith Ringle Morris, and Paul N Bennett, explores existing practices, expectations and preferences surrounding the use of imprecise temporal expressions (ITEs) with virtual assistants. Modern virtual assistants may be confused by inexact commands such as “Remind me to get milk later this afternoon.” The researchers found that people frequently use a diverse set of ITEs in both communication and planning which can result in uncertainty or improper task priority. The paper will be presented on Monday, May 8.

Also receiving an honorable mention is “Community Commerce: Facilitating Trust in Mom-to-Mom Sale Group on Facebook,” authored by PhD student Carol Moser, Professor Paul Resnick and Assistant Professor Sarita Schoenebeck. Consumers are turning to Facebook Groups to buy and sell with strangers in their local communities. This trend is counter-intuitive given Facebook’s lack of conventional e-commerce features, such as sophisticated search engines and reputation systems. The researchers interviewed 18 members of two Mom-to-Mom Facebook sale groups. Despite a lack of commerce tools, members perceived sale groups as an easy-to-use way to quickly, conveniently and safely buy and sell. The paper discusses how community commerce affords unique and sometimes superior trust assurances and proposes design implications for platforms hoping to foster trust between members who buy, sell or share amongst themselves. The paper will be presented on Wednesday, May 10.

Also presented on May 10 will be the Honorable Mention paper “Informality and Invisibility – Traditional Technologies as Tools for Collaboration in an Informal Market,” authored by doctoral student Priyank Chandra. This paper explores how actors in local markets in the Global South adapt traditional communication technologies to successfully collaborate to sustain the markets and their business practices. Drawing on observations at a local technology goods market in Bangalore, India, the study details the use of a landline telephone intercom system as the primary tool for business communication in the market. Through analyzing how the intercom system relates to informality and physical space, the paper argues that it bridges the formal with the informal, and helps facilitate informal business practices while also allowing them to remain hidden from the formal regulatory gaze of the state.

Other papers being presented by UMSI faculty and doctoral students over the six-day conference cover topics such as patient behavior during emergency room visits, personalization of news article content, daddy blogs, preferences for parents sharing information about children on social media, e-commerce choice and social mobility via ride-sharing services.

Uncovering the Values and Constraints of Real-time Ride Sharing for Low-resource Populations” by Tawanna Dillahunt, Vaishnav Kameswaran, Linfeng Li, Tanya Rosenblat.
Real-time ridesharing services such as Uber and Lyft are often touted as sharing-economy leaders that dramatically lower the cost of transportation. But how to make these services work better among low-income and transportation-scarce households, how these individuals experience these services, and whether they encounter barriers in enlisting these services is unknown. To address these questions, the researchers recruited 13 low-income individuals living in transportation-scarce environments to use Uber as passengers. The participants found these services to be reliable and benefited from rich social interactions with drivers; however, barriers such as cost, limited payment methods, and low digital literacy can make such services infeasible. The study includes suggestions that could lead to increased digital literacy and application transparency. The paper will be presented on Tuesday, May 9.

No Such Thing as Too Much Chocolate: Evidence Against Choice Overload in E-Commerce,” by Carol Moser, Chanda Phelan, Paul Resnick, Sarita Schoenebeck and Katharina Reinecke.
E-commerce designers must decide how many products to display at one time. Choice overload research has demonstrated the surprising finding that more choice is not necessarily better—having more choices may result in lower levels of satisfaction. This research tests the choice overload effect in an e-commerce context and explores how the choice overload effect is influenced by an individual’s tendency to make the best possible choice (maximizer) or be satisfied their decision even if it isn’t the best (satisficer). Consistent with prior work, the research found that maximizers are less satisfied with their product choice than satisficers. However, it also found that the number of choices didn’t affect satisfaction by much, if at all. The study discusses why the decision-making process may be different in e-commerce contexts than the physical settings used in previous choice overload experiments.

The paper will be presented on Wednesday, May 10.

For a complete list of University of Michigan participation at CHI, visit the Michigan@CHI page

Posted May 3, 2017