The study is a wee bit too small to be submitted to a conference paper, but too large to be completely forgotten, so I’m posting it online. The study tried to study the occurrence of choice overload in a movie recommender system and whether or not diversifying the set of recommendations alleviates choice overload. The results were not really conclusive, but good enough to base a second study on. And enough questions rose to base a master thesis research on. So go ahead and read it if you want to.

The Computer Vision course is hella fun! We’re working on an algorithm that stabilizes video material which it is doing pretty well already. I can’t show you any results yet, since my eee doesn’t do coding video that well . I have a small clip that shows you how we find the movement in video via which we try and remove the shaking that people usually do. Some more than others…
The video looked better before uploading to Youtube… what you should see is a movie clip with in it arrows pointing in the direction in which points in the movie move. For people that are interested: Shi and Tomasi feature finding and pyramidical Lucas Kanade optical flow. YEEEE!

And as far as holidays go, we have it locked down. The mrs. is coming over 5th of december from which we will spend a whopping 5 days in Singapore. Then fly to Krabi, Thailand for a week of sunbathing, seeing … whatever… and doing… whatever… I don’t know about Thailand. And after that we’re flying to Jogyakarta to meet a friend of ours and a teacher of mine. It’s going to be weird, yet awesome. It is kind of weird, but he is the only 40+ person who actually commented on this blog, so that kind of levels stuff out.

Anyways, now off for dinner, than bed, then school again. Cheers!

CS4243 – Computer Vision and Pattern Recognition
A course in which you learn a lot about using Matlab to analyze images. There’s a project that’s part of the course in which you are supposed to create a system which does something … visual. Examples of last year are a chess computer that uses camera’s to read the chessboard and a camera that can track moving faces in its image. I’m quite excited about it, mainly because I don’t know too much about programming. I know a lot about perception and related subjects like optical flow. And some stuff about signal analysis and using Fourier transforms to analyze these. Ah well, we’ll see how it turns out…

CS5340 – Uncertainty Modelling in AI
Presumably the hardest course I’m taking. It mainly aims at using Bayesian networks in order to model uncertainty. Bayesian networks are mathematical models that can be used to model a number of things. For the people that have been watching House (just watched s2e24, and it’s awesome!), a differential diagnosis is basically a Bayesian network. You use a number of evidences to try and find out what the most likely cause is.
An example from cognition is the word recognition model by McClelland and Rummelhart (1981) in which different aspects of words, letters and elements of letters are used to find the most likely word that is depicted by a number of letters. However, this last example might be a neural network… I’m not entirely sure yet. But I’m sure I’ll know the difference in a matter of weeks!

PL5221 – Analyzing Psychological Data using General Linear Models
Even though I already have a pretty decent background in statistics, I think it’s worth just to know a little bit more. I’ve seen a couple of things on GLM’s, but every time I need to run one on data, I encounter words I don’t know the meaning of in SPSS. So this course is supposedly going to help me in becoming a better data analyst, which will allow me to perform even more complicated experiments. It’s not like chi square will bring you anywhere right?

EE4305 – Introduction to Fuzzy/Neural Networks
The original idea was using this course to make the computer vision course easier. But they’re at the same time, so maybe it will be helpful when I get a bit further into the course. The best thing about this course till this far is the teacher. A really helpful guy, who speaks English really good… grammar wise. His pronunciation makes ‘variable’ sound like ‘rawrble’. But I don’t foresee many problems during this course.

Traveling

[singlepic id=43 w=320 h=240 float=left]Tutorials and projects are only starting next week, so the time I spend studying is limited to the lectures I attend and the preparation I do for these lectures. I’ve had only one assignment so far, which worked out pretty ok. So what do you do when you’re in the hub of South-East Asia? You travel!

Last weekend I’ve been to Pulau Bintan in Indonesia with a couple of people I met here in Singapore. A nice mixed bunch of people. Americans, Canadians, French, Germans, Scandi’s and Dutchies all getting along just fine. We rented two huts in a place called “Shady Shacks” for really no money at all. The tickets we booked were pretty cheap too. And the entire weekend was really good.

We left on Saturday too early to be healthy and at arrival on Bintan, after about 2 hours on a ferry, we stepped ashore into the pouring rain. A bit weird, a bunch of westerners walking around in the rain looking for a place to eat. It was national day in Bintan, celebrating the liberation from the Dutch reign (at the point I found out I took of the Dutch flag draped around my shoulders), so it was really crowded on the streets. Weirdest thing would have been being flipped off by a 13 year old boy… I just assume that’s the normal way of greeting people on Bintan.

[singlepic id=41 w=320 h=240 float=left]We had an awesome dinner in the harbourtown/capital Tanjung Pinang, after which we got on 2 cabs to go to the Shady Shacks, but not before going to the supermarket for some alcohol preparation. A crate of Bintang and one of Carlsberg were supposed to keep us alive and amused throughout the rainy day. And it really didn’t stop raining, even when we got to our accomodation. But since it was still warm, we decided to bring some beers into the sea and just go around for a swim in the rain anyways. Which proved to be an awesome preparation for the dinner and after-dinner.

I have to admit I don’t remember too much of that evening, but one absolutely weird incident I have to share. While I was sitting and enjoying a beer, Chris, a friend of mine came to me with a quite surprised look on his face and asked me to come with him. Apparently this local guy called Smiley was able to read hands and found out something quite personal about Chris. So ye, scientifically educated as I am, I let him read my hand and I’m still amazed by what he concluded. “When you were young your parents made you learn how to fight”, which is absolutely true… I even remember the day my mother called a Pencak Silat school in Zwolle to have me train there, where I stuck for 8 years. Black magic apparently still exists in Indonesia. Someone care to provide me a plausible reason for this guy knowing this?

At least the next day the weather was better, which leaded to us spending the day at the beach, getting sunburns and having dinner and really not doing anything worth mentioning. Or taking any pictures worth mentioning. In retrospect, that dinner gave Sam and me a foodpoisoning, but damn those shrimps were good.

Coming weeks

I’ve decided to do less partying during the week and pay more attention to school. So this is where my rhythm is supposed to be kicking in. From now on it will be all studying and spending my time wisely. There’s still a trip to Pulau Tioman in Malaysia being planned in a few weeks. Next Saturday I’m probably going to the DMC championships.

And in the recess week, starting 22nd of September, I’ll maybe go to Thailand or Bali/Lombok… only time will tell. And for the rest, I got word from Eindhoven our weekly Intermania meetings started again. Which means that I’ll be writing a small article on how Singaporeans do not understand a damn thing of environmental psychology!

_{en ettor, alvast een haeahahehehaeaha op de spitsvondige comment die je gaat plaatsen!}

Our teamname was the only good thing about the quiz to be honest

I want to forget about it, but I also want to write about it. We played with our Intermania bunch and I really expected us to do well. I mean, I had the idea we would know a lot of stuff. But we finished second last and didn’t even correctly spell ‘applaudisseren’, which is quite shameful for people that make a magazine. And we didn’t know a shitload of stuff. I mean, it’s supposed to be trivia in a quiz like that, so people with no lives that press random article on wikipedia should do good, but I kinda had the idea we were people that do exactly that. But yeah, we didn’t. A hard lesson in modesty, something that might be good for us.

 

Ah well, one of my illusions just got destroyed… Or maybe we were just smarter than that pubquiz! At least we had the best name…

Bart invited me to go to Wageningen with him, to visit the restaurant of the future. It was supposed to be a state of the art laboratory, and a cooperation between Wageningen University and Noldus, a company that designs solutions for behavioral research. I’ve had a one day training with Noldus software at the beginning of this year, so I kinda knew what they did. And I expected seeing a behavioral research laboratory in a restaurant setting to be worth my time.
Anyways, really early in the morning, going to a seminar is not something I do often, so I did not have the best preparation. After 2 hours by train and bus, I realized that going to something like this in Dickies and an Adidas jacket might not be the best way to present yourself. But feck it right? We arrived at Noldus, drank a cup of coffee and chatted with some people, which was already quite nice. But of course they had a program for us.

The day started with some presentations on Noldus. Later on people presented some use cases, which to me all seemed kinda… dodgy. In my master’s I’m being taught how to properly design an experiment. We do a literature study, in which we try to find out what factors might be of influence on what we’re trying to research, after which we design an experiment in which we try to manipulate as few things as possible and use statistics in order to say if there really is an effect. Completely unlike how the use cases were, where people just gather as much data as possible and use data mining to find possible correlations… One might call it arrogance, but at HTI we really learn how to do proper research.

One thing that was (and still is) bothering me, is the fact that one study was performed by another department of my own university. They have created a laboratory where they do exactly the same research as… our own department. Similar methods, similar research, different labs. Waste of money? Methinks so…

[singlepic id=99 w=320 h=240 float=right]But ok, after the use cases, we were driven to the Restaurant of the Future, where we had a really great lunch (or brunch actually). We had a chat with some employees of Noldus and of course tried to find all cameras used in the observations. Apparently all the cameras and other sensors produce around 60 gigs of data per hour, which has to be processed by (presumably) students afterwards. Via the control room, a number of things can be manipulated. These include smell, temperature, lights, colour, the food, general layout, and more. Things that they want to research with this data are for example the effect that the colour of the lights has on preferences for food. The restaurant is supposed to be running for 10 years, in order to have produced enough data… which is quite long…

[singlepic id=101 w=320 h=240 float=right]After this small demo of the restaurant of the future, we headed back to Noldus where we had some demonstrations of some of the new machinery Noldus has. Among others, these are eyetrackers and once again I got the impression people use these to just have some more data. Researchers just don’t know what they’re looking for anymore, which is a shame…

The whole Noldus thing coming to an end, I walked to the place of a friend of mine, with whom I had dinner. The journey back to Eindhoven was horrible, with missing trains and all. Which lead me to come back home at around midnight, 17 hours after leaving home. Talk about a long day…

Lectures and courses
I’ve been following so many lectures on so many papers. It’s been crazy. Subjects ranged from “A comparison between Heuristics Evalutation Methods” to “User Research using Mechanical Turk”. I’ve been trying to attend to a lot of usability lectures, some of which were really interesting and some of which were plain bad. It was nice to see the difference between academic work and corporate work.
On the general however, it’s just like school. A lot of people sitting and listening to a lecture. Some people taking notes on paper, some taking notes on their macbook pros and some just taking pictures of the projected slides.
I took a course on Wednesday. I figured that if I want to know anything about usability and study this in a psychological fashion, I need to know how to apply it to designs. So I registered for this course by Mike Atwood and Tom Hewett from Drexel University called “A practical approach to interactive system design”. I learned a couple of methods that serve as a basis for designing and evaluating interactive systems. I did this together with two PhD students, Malte and Javed. It was really fun, since when we were putting our skills to the test in the assignments, we did a great job. And we made puns, like ’snackquisition’, which I still like to be honest. I’m writing it down, knowing that this is typically a “you had to be there”-kinda thing, but I don’t care too much bout that now.

Social stuff
This whole course thing also lead to me going to the google hospitality event. Which basically is a free drink at some expensive ‘Grand Hotel Firenze’ aka standing in line for your drinks for too long with too many people in a place that is too posh. It was however nice to meet some people and talk about geeky stuff. It’s not that different from a gaming weekend to be honest.
I didn’t go to bed too late, since I had to present my poster on Thursday. I was scared that people would come up with some tough questions (which they did), so I decided I needed a good night’s sleep.
Poster presentation
A large part of CHI consists of poster presentations. On Wednesday Brian and Daan presented their researched. I presented a poster I made on a research on Thursday. It was quite nice, because some people even showed an interest in our research, even though it’s nothing applied like the majority of CHI papers. Also funny that most people would read the poster and then touch my upper arm briefly to ask me for comments. (check the poster if you want to know why that’s funny).

The trip back however was hell. Our plane left at 10am, which forced us to be on the airport at around 8am, which forced us to grab a train at 6:30am, which forced us to leave the appartment at around 6am, which forced us to wake up at around 5am. Luckily I did wake up early enough every day, to not be completely deregulated timewise. Anyway, after a week of Florence, Eindhoven is even uglier than it was before and colder too. It’s been a great experience to be at a conference. I did meet a lot of new people and saw people I hadn’t seen in a long time.

I had an extensive preparation prior to getting in Florence that consisted of booking a flight, getting accomodation and partying the night before departure with crazy exchange students in Eindhoven. Needless to say, the first day in Florence was hectic.

[singlepic id=56 w=320 h=240 float=left]I did make it to the apartment yesterday quite effortlesly (somewhat to my surprise). I did however completely misunderstood Karolien’s Belgian accent for Italian, which lead to me speaking English to someone who’s talking Dutch to me. And while writing this, I realize I haven’t explained that to her yet… But the apartment is way too nice for a shitty student like myself. I’m there together with 3 phd students from my faculty, who I don’t know that well. I think that this might be different when the week is over. I’ve got my own bedroom, with nice kingsize bed. And I’ve got my own bathroom, with bidet. We’re sharing a living room and kitchen, together with the conversations and thoughts that come naturally.

Day one

[singlepic id=58 w=320 h=240 float=left][singlepic id=60 w=320 h=240 float=left]The first day, due to my lack of preparation, consisted mainly of walking to the conference center (in a medieval fort) and banks close to the center. I didn’t have a registration yet, and apparently you can only pay with US currency or credit card. Having neither, I had to make a stop at the bank and having the memory and foresight of someone that woke up too early after a day of travelling, I didn’t change enough the first time. So all in all, I’ve seen plenty of the city already. Considering the weather and the absolute gorgeousness of Florence, not a bad thing. Considering this is the world’s way of telling me I need a creditcard, I’m still ok without one.
Having taken care of my registration it was time to decide on what to do this week. There’s so much to learn at a conference, that this was in fact a bit scary. I’ve found a lot of presentations on usability and registered for a basic Human Centered Design course on Wednesday. I think this might be a nice (applied) extra to my somewhat theoretical masters program. On Tuesday I’m meeting with some people from Carnegie Mellon, with whom I hope to be able to arrange something for the exchange program of my curriculum. On Thursday I’m presenting a poster (link) on one of the researches I finished last year.

And throughout the week I’ll be walking around and taking pictures of this town with my new Nikon.

Click here for the rest of my pictures.