Pervasive 2007 Conference trip report in IEEE Pervasive Computing magazine

Just found out that the Pervasive 2007 conference trip report which I helped co-author is now out in the IEEE Pervasive Computing magazine Vol. 6 No. 4 (October - December 2007). You can read the article here.

On Technorati: Pervasive2007

Doctoral colloquium slides, podcast and paper from Pervasive

I've just uploaded my slides, talk and paper from the doctoral colloquium talk from the Pervasive conference last week. Thanks to all for the great feedback. Pervasive was a great conference, and I really enjoyed meeting new and familiar faces, and explaining about my research. It was also great to be a student volunteer and helping out in whatever way I could. Next year's Pervasive is going to be Australia!

If you haven't already, check out my Pervasive photos on Flickr, or all the Pervasive photos on Flickr. At the same time, I've also updated my publications list and my web site so go and check that out.

And for those in Canada, have a happy and safe Victoria Day! For those in the US, have a happy and safe Memorial Day!

On Technorati: Pervasive07, Doctoral Colloquium

Pervasive conference, tutorial on Context-Aware Computing

The last tutorial of the conference is Context-aware computing presented by Anind Dey of Carnegie Mellon University. Context is all the stuff that is important for understanding a user's goals and motivation. What is the important part for context is reminders, when and where to deliver context? It is easy to grab context, but more difficult in my opinion, to infer context. Context is a very difficult and slippery word to define because it means different things to different people. Inferring an activity or task as context is difficult. Finding social norms as context is, according to Anind, is slippery. My PhD work is looking at community as a metaphor for finding social context. I believe that there is some way that we can infer social context automatically from the links within the social hypertext. As a pervasive computing community, we need to look at real problems instead of recycling over the same applications like location guides. Anind says that grad students should not do work in context infrastructures because it is extremely difficult to publish your work.

One of the challenges of context is how to leverage real human context and realize that you're not really getting at human intention. Anind showed some videos of examples of context, one video was about an automatic door made in Japan that conforms to the type of body you have and will open that much.

Pervasive conference, tutorial on Machine Learning for Ubicomp

Bernt is presenting the tutorial on machine learning for ubicomp. Machine learning comes from artificial intelligence and I think this is needed to make ubicomp systems intelligent and more smart. How to learn to perform a task from experience? Learning has a lot to do with optimization. We can use Bayes decision theory to classify some class based on some previous class. We can also use naive Bayes classifier, we need the probability given a class, and the probability of the class. There are graphical models of Bayesian networks such as Hidden Markov Models (directed, dynamic Bayesian network).

Pervasive conference, tutorial on Evaluation of Real Deployments in Ubicomp

Gregory Abowd from Georgia Tech is giving the tutorial on Evaluation of Real Deployments in Ubicomp. He is talking about what are the issues you need to think about when deploying a ubicomp system. Very often papers get rejected because there is no evaluation of the system in a deployed environment. What is important is how to do research with human subjects, how there needs to be an ethics approval, compensation, getting the right demographic and retaining participants. Sometimes it is OK to be the end-user and the designer. We also have to ask users what they think through surveys, questionnaires, and interviews. You need to spend a lot of time to prepare the survey or interview, but in the end the analysis is much easier, so basically you should not rush writing a survey or questionnaire.

In designing a study, you need to be a devil's advocate and ask yourself what things could affect the interpretation of the study and controlling variables. What kind of study is it, within subjects design or between subjects design. How do we get information from a person's every day life? We don't want to follow them every hour, every minute, by instrumenting a camera or recording device. Instead, the better way is to use experience sampling, where you interrupt the person to gather relevant information.

With experience sampling, you can use a diary where you ask the participant to record what they were doing, where they were doing it, why they were doing it. However, there are social-legal implications of recording this. Basically, in the end, we need to balance tracking participants with asking participants questions so that can be able to match between them.

Pervasive conference, tutorial on Ethnography

This next tutorial is on Ethnography: Thick and Thin being presented by Ken Anderson of Intel and Paul Dourish of UC Irvine. According to Marilyn Strathern, Cambridge, ethnography is "the deliberate attempt to generate more data than the investigator is aware of at the time of collection". Ethnography has started since the 1910s, even further than computer science, with anthropology. People think of ethnography as a tool for studying long-term studies and participatory observation. Ethnography is about not just about capturing the experience, but taking that experience, and determining how this experience can relate to the real world. So, ethnography is not about replicating the experience, making you become part of the culture that you're studying. Tom's talk talked about just because you can do this, doesn't mean that you should, whereas with this talk, just because you should doesn't mean that you can which is the opposite.

Ethnographic data is interpreted, and I believe that culture is important in understanding, deploying and evaluating ubiquitous computing systems. When you deploy a ubicomp system, how is this going to affect the culture? If it starts breaking culture, people won't use it. The systems need to be embedded within the culture, without disrupting it. Ethnography is about writing about people, after the things happen not during. The ethnographer has two roles: one as author and one as participant. So what does ethnography have to do within ubiquitous computing, we can think of technology as a site of social and cultural production. What are the good questions to ask of an ethnography? What are the empirical claims? What theoretical claims are made? How does this contribute to the corpus? What was the context of production?

Pervasive conference, tutorial on Interaction in Ubiquitous Computing

Right now is the tutorial on interaction and ubiquitous computing presented by Tom Rodden of the University of Nottingham. Tom is talking about how do we achieve Mark Weiser's vision? Technology is vanishing into smaller and smaller devices which we don't now even think about, like USB keys and iPods. What are the key issues of ubiquitous computing? First is natural interfaces and an example of this is Stanford's iRoom which is a smart room environment where there are wall-mounted displays and controllers are used to transfer files to the display and to interact with the tabletop display. The real challenge for HCI is how may these things be used in real life. The second part is context, how to use context to provide awareness? We need to ask these questions when we measure and use context: who, what, where, when and why. The challenges of context-awareness to add more context or to fuse data to infer context.
Digital information is being augmented in the physical world, actions on physical objects have meaning in the virtual world and vice versa.

One of the issues that start to emerge is that will each object have a little instruction manual? One of the general theme of this tutorial is that we design interaction in ubicomp, but will they actually be used in everyday life. Tom showed some really neat videos, one with a weight table and key table, that based on weight, the image moves faster or the picture tilts more. Attendees in the audience laughed, because would you have that kind of device in your home? Nonetheless, a great talk, I thoroughly enjoyed it.

Pervasive conference, Day 4

Today is Day 4 and the last day of the Pervasive conference, with the tutorial sessions. I'm once again doing my duty as a student volunteer, so if you see me in between breaks, come and say hi! I'm really enjoying this conference, it was a great chance to meet lots of new people, researchers that I've always wanted to meet. Dinner last night at the CN Tower was great, I really liked the appetizers that were given to us while we were in the cafe. I've been to the CN Tower many times, but never to the cafe, and only maybe twice at the 360 revolving restaurant. I took some pictures, although others at the conference on Flickr will have done a much better job than my Palm Zire 71, so you can check them there.

There will be a student volunteer party tonight for all the hard work we the student volunteers did!

Pervasive conference, Day 2

Today is Day 2 of the Pervasive conference. Adam Greenfield is the keynote speaker and will talk about Everyware. Khai is the conference chair and he is now introducing the Pervasive conference and talking about why Toronto is a perfect fit for hosting Pervasive 2007.



This is the first time Pervasive has ventured outside Europe, there are over 300 attendees from Europe, Asia, Australia and North America. Khai is talking about the structure and composition of the Pervasive conference. There will be posters and demo presentations today, along with papers. Khai is thanking all the various people that made the conference possible, including the student volunteers yeah! And now, Khai is thanking the industry sponsors. All photos from the conference will be tagged in Flickr with the tag "pervasive07".

This year, there were 142 papers of which 132 were valid. There were 403 submitting authors from 32 countries out of which 21 papers were accepted for a 16% acceptance rate. For the paper review process, all papers were given to the two PC members, where there were a total of 4 reviews per paper. There were 28 PC members, 178 external reviewers, and 532 completed reviews. Khai is introducing Adam Greenfield and how his work is considered as "soulful" by Bruce Sterling, the keynote speaker at Ubicomp 2006.



Adam had a 200 slide Flickr presentation at the CHI conference. Adam's book on Everyware is the first book on ubiquitous computing that the lay person can understand, trying to make sense of Mark Weiser's vision. Some characteristics of Weiser's ubicomp are embedded, wireless, imperceptible, multiple, post-GUI. Adam says that Weiser's vision is considered still radical, and a lot of the things PARC had to create from scratch, which we take for granted right now like wireless 802.11. We are now beginning to see an emergent "internet of things", a class of systems that tends to colonize everyday life. This is an aggressive claim over a large domain. Adam says that "information processing" is "dissolving in behavior". One example of this that Adam gives is about the Octopus card in Hong Kong used for transit which is a wireless reader card where you can wave it towards the reader without having to take out the card (women in Hong Kong leave that card into their handbags). So what happens is that the visible (eg. the card) becomes invisible. I can say that I really love the Octopus card, Toronto transit should adopt that (so I don't have to use tokens for the TTC and GO Train ticket for the GO Train, and I can use the same card to take the ferry, the Oakville bus, York region transit, or any other transit system in the GTA).

So is this science fiction? Adam believes that we are beginning to see some of these visions happening but in a different fashion. The opt out feature is probably something that ubicomp researchers don't really think about when designing ubicomp systems, so to me that is a new thing. I met Adam before in San Francisco when I was doing my internship at PARC and had the privilege of doing a small interview with him which you can see from this post on my blog.

I just uploaded my photos that I took from my Palm Zire 71 up on the Flickr site, tagged with pervasive07.

On Technorati: pervasive07

At Pervasive today, workshops and doctoral colloquium

The Pervasive conference starts today with the workshops and doctoral colloquium. As a student volunteer, I helped with stuffing conference bags with proceedings and adjunct proceedings, and giving them to attendees. I'm right now in the student volunteer room now reviewing my slides for the doctoral colloquium and hanging out with other student volunteers. I also took pictures from the conference so far which will be uploaded later.

On Technorati: Pervasive07

Pervasive conference starts tomorrow!

I'm going to give a doctoral colloquium talk tomorrow at Pervasive called "Searching for Sense of Virtual Community in Social Hypertext". Looking forward to some great discussion. I'll also be working as a student volunteer as well. If you're at Pervasive, come and see me, I'll be working at the registration desk tomorrow from 8 to 10 am. Then, my talk is at 2:30 pm.

Got to finish those doctoral colloquium slides!

On Technorati: Pervasive2007, doctoral colloquium

Presenting and volunteering at Pervasive 2007

I will be at the Pervasive 2007 conference which is in Toronto this year. Finally, I've always wanted to go to Pervasive but never had the chance to go to because it was too far away and too expensive. I will be presenting my doctoral research as part of the Doctoral Colloquium which is "The Search for Sense of Virtual Community in Social Hypertext". As well, I will also be a student volunteer at Pervasive, so if you're at Pervasive, you might see me running around the conference, or helping out at different sessions, or even socializing. Pervasive starts May 13 and runs to May 16. I'm really eager to help out and to be at this conference.

The keynote speaker is Adam Greenfield, who is the author of the Everyware: The Dawning Age of Ubiquitous Computing book. Adam will be talking about "Everyware: Some Social and Ethical Implications of Ubiquitous Computing". That's great, because I met Adam when he gave a talk in San Francisco last summer when I was working at PARC, you can see my blog post about him here.

Pervasive should be a great conference, and I'm looking forward to it!

On Technorati: Pervasive2007, Toronto, Ubiquitous Computing, pervasive computing, doctoral colloquium