QS Public Health Symposium



Last week I had the opportunity to attend the Quantified-Self Public Health Symposium organized at UCSD by Bryan Sivak (U.S. Department of Health and Human Services), Larry Smarr (Calit2 and ), and Gary Wolf (Quantified Self Labs). The intent of the one-day event was to provide a conducive environment for researchers, policy makers and developers of QS-type tools to discuss issues around data and its potential for use in public health. Many attendees such as Susannah Fox have written about the highlights of the day from their perspective. Here are my notes:

At the beginning of the day, Larry Smarr said he believes this is the “year of scale” when it comes to personal health informatics. Stephen Downs (Robert Wood Foundation) reiterated this point with a remark about our “rapidly expanding oceans of personal health data” and expanded on how RWJ Foundation has identified that a cultural change is needed around the value of health in society. Gary Wolf pointed out the uniqueness of the researchers involved in QS, ranging from public health epidemiologists and CS/HCI types, and also claimed that we are at “the beginning of a new program in the human sciences”, which I very much agree. Susannah Fox from the Pew Research Center made a number of excellent points based on the data she has been collecting and studying. When it comes to self-quantification, do “people want the unforgiving bright light of numbers in front of them”?, she asked. Fox also talked about how it is important to respect cultural rituals.

Brian Sivak was one of the major forces in the event. It was a pleasure to see someone from a government agency be so forward-thinking about the opportunities in this field. He discussed a number of topics, from areas he considers critical (e.g., data ownership, privacy and standards) to the possibility of a broad public health campaign around activity tracking. He also introduced the theme of sharing documents such as terms of service and privacy policy agreements amongst tool developers to facilitate the exchange of personal health data.

Later in the day, in a session reserved for “toolmakers”, Anne Wright from Fluxtream discussed how personal health data should be as portable as personal financial data, where one can download documents from banks and upload them to tools such as Mint and Quicken. In terms of visualizing and making sense of data, she mentioned the need for different types of interfaces at different levels (e.g., personal dashboards versus deep history introspection). Margareth McKenna from Runkeeper brought with her a number of research questions that are difficult to answer even in light of large amounts of data, such as what happens when people are not logging their exercise and activities (“sensors can only capture so much”). She also emphasized the need to disseminate tools to everyone, echoing others during the day who talked about the “digital health divide”. Andy Hickl from A.R.O. presented the “11-week problem”, at which point self-tracking devices go to the drawer. In other words, “what should we care?”, or how can we make self-tracking data meaningful? At the end of the session, Gary Wolf came back and proposed a research toolkit that could be shared within the community.

In one of the sessions focused on study designs, a highlight for me was Eric Hekler, a behavior psychologist and deep thinker around behavior change. He discussed how he lost faith in randomized clinical trials and how we should be asking people what they truly want from data and devices. He also made the point that we should support people to come up with their own interventions and find out ways to “get more functional life years”.

To me, the high-point of the day was a panel with Larry Smarr and Lee Hood. Hood, whose work has revolutionized biology and genetics, is leading a longitudinal, Framingham-like study with the goal of better understanding how digital technologies and ongoing self-tracking can be leveraged to quantify what it means to be healthy. Another goal of the study is to look at the progression of disease, as observed by these devices. Smarr, one of the most recognized self-trackers in the Quantified-Self movement today, started his tracking journey to understand his own personal health issues, in a way doctors could not, or were simply unwilling to listen.

In addition to the plenary talks and discussions, the day was filled with coffee breaks, where I had the chance to chat and exchange ideas with many researchers and developers, such as Kevin Patrick from UCSD and Mike Lee, who leads MyFitnessPal. Overall, a great day in the campus of UCSD.


Hello 2014

A new year is upon us, hello 2014. This is the time of the year when I look back and wonder how 12 months went by so quickly. I am sure I am not alone. The last couple of months of 2013 were particularly busy. In addition to personal trips during the holidays, I also had the opportunity to attend the 2013 SenseCam and Pervasive Imaging Conference in San Diego, CA.

This was a small conference focused primarily on applications of first-person point-of-view images. I presented our paper titled “Feasibility of Identifying Eating Moments from First-Person Images Leveraging Human Computation”, which can be downloaded here. Our work was very well received and I had the opportunity to meet several researchers at UCSD and beyond who are taking health and behavior assessment to a whole new level thanks to wearable cameras.

Next on the agenda are additional studies related to my thesis work and my thesis proposal, which will probably take place at the end of January. As @dgmacarthur wrote on Twitter:

“May 2014 bring you clean data, well-documented code, large and deeply phenotyped samples, and a clear path to clinical translation”

Video of Ubicomp 2013 Talk

Thanks to my friend and fellow graduate student Gabriel Reyes, here is my talk at Ubicomp in Zurich a few weeks ago. I presented our paper titled “Technological Approaches for Addressing Privacy Concerns When Recognizing Eating Behaviors with Wearable Cameras”. You can download the paper here. Gabriel recorded this video with his Google Glass. Headphones recommended.

Ubicomp, Doctoral School and CHI

Academically and intellectually, September was the busiest month of the year for me, and the weeks leading up to it were quite intense as well. In light of conference travels, presentations and deadlines, there weren’t too many hours left in the day. Luckily all activities were tremendously exciting.


From September 8th to the 12th, I attended the Ubicomp 2013 conference in Zurich. I participated in the Doctoral School and also presented a paper on the last day of the conference. This year ISWC (International Symposium on Wearable Computers) was co-located with Ubicomp and I think this was a great move, since there was an obvious and productive overlap between both conferences. I should also say that the Ubicomp and Pervasive conferences merged, and this year we had the first joint conference between these two events, which is now officially called the 2013 ACM International Joint Conference on Pervasive and Ubiquitous Computing. The 4-track conference was very well attended with 700 people registering. More on Ubicomp 2013 in future posts.


Once back in Atlanta from Zurich, there was nothing but long hours and lots of work in preparation for a submission to CHI, whose Paper & Notes deadline was September 18th. At the end of the day we felt that we submitted a strong paper to be considered for the CHI 2014 program.

ISTC-PC Summer 2013 Retreat

One of the sponsors of my research at Georgia Tech is Intel. My group, the Ubicomp group, is a member of the Intel Science and Technology Center for Pervasive Computing (ISTC-PC), and together with other academic institutions and researchers, we explore how sensing and computing technologies are changing the landscape of a variety of domains, such as health and wellbeing.

Every year, Intel organizes a retreat event with all ISTC-PC members. This year we all traveled to Suncadia, about one hour east of Seattle for 2 days of face-to-face conversation. Some of the topics covered included the future of sensing, privacy & security, activity recognition and behavior journaling.


It was a very productive event with good discussions and the opportunity to meet many friends in the research community. And the location was truly spectacular.

HCIC 2013

As I write this I am on my way back from the HCIC workshop, a summercamp-style gathering of leading HCI researchers that has been led and organized by Gary and Judy Olson for many years. One of the unique aspects of HCIC is that it is more than a conference – it is actually an organization and only invited members can attend the three-day workshop. Georgia Tech is a member, and this year I was one of the lucky few students selected to join the group at Asilomar, near Monterey, CA.


The theme this year was “data, data, data” and all the sessions were about the impact of Big Data. Privacy, theory, metrics, and visualization were just some of the topics covered. Eric Brill from eBay Research kicked off the event with a talk on how “oodles” of data allow for virtually free UX experimentation and a new way of thinking when it comes to building data models and inference algorithms. He urged us to think about ways to enrich the input “signal” beyond clicking streams since there is so much emotion and passion behind user actions.

There were too many highlights to count over the 3 days of the workshop. Personally, I very much enjoyed the crowdsourcing and information visualization talks. Some important privacy and ethical issues were discussed, and the role of theory in HCI was a recurring theme. Perhaps my favorite “big idea” from the workshop was the notion of combining techniques from machine learning with qualitative approaches so that we can discover not only the latent structure in the data, but the meaningful latent structure in the data. This was articulated by Jim Herbsleb and Rogerio de Paula, who pointed out Ken Anderson’s work on ethno-mining.


HCIC, thanks for the beautiful venue, intellectual insights, and quality time with new and old friends.

CHI2013 Review

It’s been about a week since I got back from Paris after attending CHI. This was the most productive CHI for me personally, which I attribute in large part to the Personal Informatics workshop I attended.


The workshop, which took place on Saturday and Sunday before the conference began, was an opportunity to spend time and exchange ideas with a great group of folks. Thanks to Ian Li, Jon Froehlich, Jakob Eg Larsen, Catherine Grevet and Ernesto Ramirez for putting it together. It required lots of work to organize it, since it was set-up as a two-day hackathon. All the effort paid off nicely in my opinion. The interesting question now is how to make the Personal Informatics community within CHI grow, since it has outgrown the workshop format.

Now on to the main conference. With 10+ tracks, it is always a challenge to navigate CHI. Here’s a timelapse of my day 1 at the conference:

I was able to attend some good/interesting presentations, such as:

Validating a Mobile Phone Application for the Everyday, Unobtrusive, Objective Measurement of Sleep, Lawson et al.

Designing Mobile Health Technology for Bipolar Disorder: A Field Trial of the MONARCA System, Bardram et al.

Footprint Tracker: Supporting Diary Studies with Lifelogging, Gouveia and Karapanos.

Mind the Theoretical Gap: Interpreting, Using, and Developing Behavioral Theory in HCI Research, Hekler et al.

NailDisplay: Bringing an Always Available Visual Display to Fingertips, 
Chao-Huai Su, Liwei Chan, Chien-Ting Weng, Rong-Hao Liang,
Kai-Yin Cheng, Bing-Yu Chen

Food Practices as Situated Action: Exploring and designing for everyday food practices with households. Rob Comber, Jettie Hoonhout, Aart T van Halteren, Paula Moynihan, Patrick Olivier

Not surprisingly I was particularly drawn to the health-focused talks, but there were some intriguing and inspiring topics of interest in many other sessions.

Thanks for the inspiration Paris. Next year, Toronto.

ICWSM-13 and Health

The 7th International Conference on Weblogs and Social Media (ICWSM) is happening in Boston this year (July 8th-11th). I am not attending but I know that there is a large number of social media researchers who have been exploring the domain of health recently. In light of this, I like to check the program of ICWSM every year. Here are some of the papers that I will make sure to take a look at in July:

Predicting Depression via Social Media
Munmun De Choudhury, Michael Gamon, Scott Counts, Eric Horvitz

Perception Differences between the Depressed and Non-Depressed Users in Twitter
Minsu Park, David W. McDonald, Meeyoung Cha

Fitter with Twitter: Understanding Personal Health and Fitness Activity in Social Media
Rannie Teodoro, Mor Naaman

Characterizing Geographic Variation in Well-Being Using Tweets
H Andrew Schwartz, Johannes C. Eichstaedt, Margaret L Kern, ukasz Dziurzyski, Megha Agrawal, Gregory J. Park, Shrinidhi K Lakshmikanth, Shneha Jha, Martin E. P. Seligman, Lyle Ungar, Richard Lucas

I have to say I am very curious about the last paper on this list.

Mobile Notifications Increase (Food) Logging Behavior

It is almost time for our community’s annual CHI pilgrimage. I am very much looking forward to it, as CHI offers a wonderful and unique opportunity to see old friends and make new ones. I’ll be there starting on April 27th for the Personal Informatics workshop.

As part of my CHI preparation activities, I’ve begun taking a closer look at papers that will be presented at the conference. I will go over some of them here, as time permits. One note that caught my attention was “The Power of Mobile Notifications to Increase Wellbeing Logging Behavior”, available here. I have been exploring the domain of food logging lately and one of the challenges has to do with the amount of work people have to do to log their every eating activity.

In this paper, Bentley and Tollmar discuss results from a study where “passive mobile notifications increase logging of wellbeing data, particularly food intake, in a mobile health service. Adding notifications increased the frequency of logging from 12% in a one-month, ten-user pilot study without reminders to 63% in the full 60-user study with reminders included”.

First of all, it is great to see studies with such a large number of participants. This is something I am trying to aim for in my own research. It looks like the second study discussed in the paper lasted for 3 months (summer of 2012), which is also a nice time frame for a study of this nature. In terms of the findings, I am surprised by how much the addition of a small notification icon in the UI and the ability to customize the notification improved logging. It is also interesting to note that even after a 5x improvement in logging frequency, users averaged only 9 days of logging for both food and weight.

Yes, we have a long way to go before food logging can be done as easily as with other forms of activity logging. Based on my recent research, I suspect the sweet spot will involve a combination of manual and automatic logging, but more validating work needs to be done, without a doubt.

Asking Good Questions at Conferences

When you attend an academic presentation at a conference, are you inclined to ask questions when Q&A begins? Or are you the type who prefers to sit back and take it all in, including the Q&A session itself? One of the things I noted at the Ubicomp conference in Pittsburgh last month was that while some of the presentations generated lots of discussions, with many people stepping up to the microphones to ask questions, others didn’t. Why is that the case?

My advisor, Gregory Abowd, is one of the senior members of the community and always likes to ask questions after presentations. In fact, he is known as someone who asks great questions. Right after the conference, a discussion within the steering committee ensued about how other senior members of the community should follow Gregory’s lead and ask questions as well. Gregory was encouraged to comment on his “secret” for asking good questions and having the attitude to do so. I reproduce here what he wrote, with his permission.

I am reminded of a comment that was sent to me secondhand after the HUC 2K conference in Bristol. A student commented to a student of mine that I had an unfair advantage at these conferences because I had already read all of the papers in advance as part of the program committee and had formulated my questions for the author well in advance. Interesting perspective, and amusing for me to think that I had enough time in my life to read all of these papers and generate questions.

As you should know, I can’t stand when there is no discussion after a paper, so if nobody else has a question, I always want to jump up and ask something. But James’ observation is not so much about whether questions are asked but the form and content of those questions and how they may or may not provoke deeper discussion. I appreciate that he feels I have value added in asking the kinds of questions I ask. I am sure not everyone shares James’ opinion on how I conduct myself at these conferences, but to the extent that people want to do some of the things I do, I will respond to his request.

What’s my secret? Let me ponder that in this response. This is stream of consciousness and in no particular order.

1) I pay attention to the presentations. How many people are at these conferences with their laptops/phones/tablets open doing other things? I am not saying that having the technology there is a bad thing. In fact, I love having the proceedings available in digital format, because I often find myself coming up with a question and wanting to see how that is addressed in the paper itself (so that I don’t have to ask). But first and foremost, I don’t think you can engage in a discussion or ask a reasonable question if you are not paying attention. Ubicomp is one of the few meetings I go to where I actually get anxious during the breaks if my conversation with people in the halls is preventing me from seeing a paper presentation. Networking with others at Ubicomp is not nearly as important to me as hearing the presentations and thinking about them.

2) A good paper presentation makes someone want to read the paper, but we don’t all have time to read everything, so the presentation discussion time is the best time to get information that you might get from reading the paper but know you won’t ever do. Q&A helps me form an opinion about the potential impact of someone’s work. I really enjoyed the discussion of a particular best paper nominee at the conference because it revealed to me that the paper was not nearly as impactful as it might have appeared from first reading (the questions were not offered by me, but they were spot on to the crux of the paper and the presenter gave very honest answers that helped us to understand limitations of the work that might not have been obvious from the presentation). So, when you are listening to a presentation, think to yourself why would you or anyone else want to build off of this work and what would be the concerns or limitations or advantages.

3) Papers are not presented in a vacuum. Don’t think about the paper presentation just on its own, but on how it relates to anything else we already know. I often think of related work during a talk and quickly scan the list of references for the paper to see if the authors mentioned it. If not, then it is the basis for a question. But also, the program chairs often spend a lot of time putting papers together in a session that have some common link. Use the Q&A to help draw out those commonalities, or contrasts. There was a great example of that this year in a session that had three papers that were very human-centric and application driven relating to kitchens and closets and a closing paper that was strictly about using RGB-D cameras to detect actions, with the example being cooking. What a wonderful opportunity for us to explore the gap between the application driven research and the technology driven research. I was so energized by that contrast that I whipped out my laptop and started sketching some ideas for a paper on how we are to advance as a community by bringing together these contrasting views on the same application space (the kitchen) and understand what the gap is between what technology can deliver today and what it needs to deliver tomorrow. I am interested in how we can systematically address that gap, but at a minimum we can have discussions about the gap when it is presented to us in the same session at a conference.

4) We are a community and the more we get to know each other, the more we should begin to behave like caring friends and family. By that I mean we should be able to speak frankly and constructively with each other in an attempt to make each of us better at what we do. I am often brutally honest with my collaborators at Georgia Tech, and the excuse I use is that I would much rather they be beaten down within the family of Georgia Tech, and then built up so that when they are outside GT they have a much better way to present and defend their work. But the key there is that you have to help someone build back up ideas that you might shoot down. I often am very good at shooting down and not so good at helping to build back up. If we could develop more of a feeling within Ubicomp that we are not competitors as much as we are collaborators seeking some common goals, and that someone else’s success is somehow our own success, then I think you would see more people trying to make others’ work better through discussion. That’s easy to say, and hard to follow.

5) Don’t be afraid to be wrong. I will bet that half of the questions I ask contain some evidence of me not understanding something or just plain being incorrect in my assessment. That’s OK. When I ask a question, it’s not about having someone else think I am smart. It is about trying to make me smarter about something I might not completely understand. Then again, there are times when I am pretty sure I am right and I am trying to point out some other way to consider the work that has been presented. When I ask a question or make a comment, I should be prepared to defend my position or my question, as much as I am putting the presenter on the spot.

6) Why do we assume that during a Q&A session, the questions come from the audience and the presenter is the only one who can answer? Let’s treat the setting more as a discussion. If you were at the paper I presented, it was wonderful to have so many people step up to comment and question my statements. But the problem was, in my opinion, I felt obliged to respond to every single comment and question, which meant that I spoke too often. I like it when the audience discusses among themselves, privately or publicly. So I often turn to the audience to make my statement or question, or direct my question not to the speaker but to someone else.

7) How to formulate a question? I think there are lots of people in the audience who have great questions and understanding of the presentations, but for some reason they don’t share their thoughts. There are lots of reasons for this, but if everyone keeps their thoughts and questions to themselves, then we are left with situations that James described. How can we encourage more ways to generate interesting questions and discussions, during and after a talk. Well, social media seems like an obvious approach, and I am sure it is already happening a lot. Can we bring the social media up front during the discussion phase of a paper, or perhaps even at the end of the day?