Highlights from CHI 2018 Montreal
At CHI I had a phenomenal experience at a paper session that was not directly related to my research. There was a paper on how to formalize a vocabulary for smart "Internet of Things" technologies. The talk was very good but it helped me figure out a problem I have been dealing with in my research where Technology on the Trail does not have a formal way of understanding what is a "trail" or what the boundary of "outdoors" is in context. Even today it gives me a lot to reflect on. How should I steer my research with regards to careful verbiage? Or perhaps become one of the first to add some formalization of vocabulary to Technology on the Trail.
My highlight of the trip was the CHI workshop on Sensemaking in a Senseless World. Virginia Tech played a central role in this workshop and made a strong case that VT is one of the major players in this research area. There were 24 accepted papers and 5 were selected for longer spotlight presentations. Two of these 5 spotlights were from VT folks. I presented our Context Slices work on behalf of Tianyi Li, and John Wenskovitch presented his work with DAC. The workshop brought together a fascinating group of people and papers and Dan Russell, the lead organizer, is taking steps to arrange a follow-up event at Google HQ.
One of my favorite talks at CHI was by Michael Veale. The talk was entitled "Fairness and Accountability Design Needs for Algorithmic Support in High-Stakes Public Sector Decision-Making” https://dl.acm.org/citation.cfm?id=3174014), and in it Michael talked about public workers who use algorithmic systems to support decision-making. He focused the value of the algorithmic systems in the public sector, and how they’re deeply embedded in the social and organizational context that they’re deployed. One particularly interesting thought I took out of his talk is this (paraphrased): sharing machine learning models is easy and we know how to package up models, but how do you share the hard-won organizational processes that have been built up around the algorithm?
I am extremely grateful and lucky to have been selected as a Student Volunteer (SV) for CHI 2018, my first conference as a PhD student. While I helped with some of the behind-the-scenes work that usually takes place before and during CHI, I was able to interact closely with many of the two hundred SVs from universities all over the world, some of whom I hope will be long-term collaborators and friends. In addition, we were given the opportunity to have lunch with well-known researchers in our field through informal 'Rockstar Lunches' that were held each day, and I was able to speak with Meredith Ringel Morris, Steve Whittaker, and Michael Muller. As a newcomer to the CHI community, being a SV not only gave me the opportunity to connect with a diverse group of people, but also let me give back to the community. I also learned first-hand that conferences like CHI would not be possible without the thousands of people who voluntarily serve in diverse roles that range from being a student volunteer, to a reviewer, to the president of SIGCHI. Overall, I believe my experience at CHI 2018 was augmented by my role as a SV, and I hope to serve as a SV in the future.
I attended two workshops, each of which was useful and interesting, but the highlight of the CHI Conference itself was a very unlikely session: “Surveys from Foundations and Trends in HCI”. Foundations and Trends is a journal, not associated with SigCHI (or ACM, for that matter) -- and it turned out only one of the two 40 minute presentations was based on an article in the journal. The first session was Jeffery Bardzell talking about a survey article on maker culture around the world, with a strong emphasis on its diverse and contradictory meaning and the inherent tendency to move towards neoliberal appropriation. The second presentation was a presentation by Mark Blythe on how developed and remarkable design fiction is. (I wish I had his slides!) It revolved around a fictional Latvian science fiction writer, but referenced a number of prescient science fiction stories (e.g. an 1890’ss story about the sinking of the large ocean liner, the Titan that matches the sinking of the Titanic 10 years later).
There were two presentations that particularly stood out to me. The one that resonated with me most wasGender Recognition or Gender Reductionism? The Social Implications of Automatic Gender Recognition Systemsby Foad Hamidi, Morgan Klaus Scheuerman, Stacy M. Branham. Part of this could be because I there are efforts in my lab toward gender recognition of Twitter accounts, and I’ve had some thoughts about how restrictive some of the generalizations were. I was really pleased to see that researchers were making efforts to show how Automatic Gender Recognition could be hurtful to people (especially since it was driven by a Hokie alum!). So much so that this is the paper that I have been telling everyone about… even people outside of computing. Since I am sure this presentation was impactful to many others, I’ll also mention another favorite presentation of mine. I initially went to theNorms Matter: Contrasting Social Support Around Behavior Change in Online Weight Loss Communities byStevie Chancellor, Andrea Hu, Munmun De Choudhury presentation because I had stumbled upona lot of Dr. Choudhury’s papers during lit reviews for some social computing/clinical psychology papers and jumped at the opportunity to see a presentation on her work. I really appreciated how it drew attention to to the fact that some online communities can be supportive toward harmful weight loss goals. The the social norms in online health communities for healthy behavior change and harmful behavior change are actually similar, and the terms used the comments of both communities were similar. However, the context for which the terms were used was significantly different. I really appreciated this presentation for its study in social norms of communities (inline with my research interests), as well as the fact that it addressed online health communities (a rather popular topic) and showed that some online health communities can facilitate unhealthy behaviors.