Sunday, December 13, 2009

My "pervasive" experience at the PC meeting

It was the first time for me to visit the Pervasive PC meeting. It was great to participate in the discussions, see the procedures of decision, and experience the emerging group dynamics. I also admired Antonio Krueger about steeping back, consequently keeping his own opinion out of discussion, but moderating the flow of arguments effectively instead.

I was really fascinated about the entire procedure: based on the expertise of the PC members and additionally reviewers they have invited, arguments in favor and against papers had been collected. The grades provided by the reviewers provided the line of sequence in which the papers had been discussed: high-ranked papers did not create long discussions and where clearly in, the same on the bottom clearly out. However, each member could intervene, so in fact it was more about valid arguments since the numeric grades which can be kind of arbitrary.
Most of the time obviously was dedicated to these "borderline" papers. For those papers it is really that they inspire at least one PC member that defends this paper with passion throughout the discussion. This is most likely to happen if the PC member is in this very specific field and believes its importance, so picking the conference with potential champions is an important strategy when submitting papers.

My lessons are:
  1. Papers have to reach the right reviewers. Reviewers that are in the very same field and can enthusiastic about the topic. Reviewers are reached by addressing the topic of the paper well in the title and abstract.
  2. Papers have to be written in a way that the story and conclusions can be conveyed to the primary and secondary reviewer in a way such that they can present the papers in the PC meeting in 1-2 minutes. As such abstract and conclusions are the most important part of a paper.
  3. Additionally, research has to be solidly conducted such that obvious objections ("to little users in the study", "advancement from related work is unclear", "statistics cannot be comprehended") do not apply.
  4. Eventhough blind submission is required the majority of the PC members considers comprehensiveness of the paper is more important - "anynomity is a means to an end". As such authors should certainly remove obvious references to their institutions but shouldn't spend to much time on anonymizing pictures and other data relevant to the story of their paper as in the end it does not matter.
  5. Authors should not submit notes (short contributions) as reviewers do not really understand the concept of notes. Also US people don't consider this as valuable contributions as they do not count for their metrics.

I really enjoyed to experience the passion and dedication all PC members brought to the table (from my perception laptops were only used to dig up reviews and papers, no emails or other unrelated stuff) in order to really identify the best papers that move the field and the conference forward. Eventhough this might be sometimes hard to believe when receiving reject reviews about a paper oneself was really convinced during submission.

Before the meeting we had a research "speed-dating". Various people presented their work. What I found interesting was this Sesame-kid book that could sense which page. As young children don't like to answer to the phone this book could be used with two maemo devices (yes, it was presented by Nokia;) to faciliate another way of telecommunication for children: having this book at each location grandparents and children could jointely read the book, see each other in the screen, and if the communication gets stuck the toy character on the left hand-site (I found it really disturbing) brings up some laugh and smile to trigger the communication. Future probably won't happen that way, but this remote communcation is definitely an interesting problem.

Family Story Play: Reading with Young Children (and Elmo) Over a Distance from hayes Raffle on Vimeo.

Friday, December 4, 2009

Digital kills real

Today I received a quite remarkable auto-reply:
Dear email sender,
I receive on a normal day around 100 emails which takes about 4 hours to read and answer. After a trip or vacation some 1000+ emails are waiting for me. I thus decided not to read all email any longer. I apologize for any inconvenience it will cause.
If you need to contact me, please do not use the phone as this is also given me similar problems and I had to switch it off.

You can always call the secretaries to leave your tel number or email address (+xxx xxxxx xxxxx) between 6:30 and 7:30am in the morning it is awfully quite and I would like to chat with you then. Later I will be too busy.

Alternatively you may want to write me a conventional letter. I would be delighted to receive one once in a while and probably even read it.

best regards


Poor guy, wasn't IT meant to make life more efficient and more effective? Now it's starting to absorb us. This is not really what I mean when aiming at "integrating the real world with the virtual world". But it's probably get worse. How can we cope with all that emails, tweets, blogs, facebook status updates? Technology can help to interweave places, things, people and information but there always only one timeline pressing ahead quickly.
However, I'm not convinced that this radical opt-out as described above should prevail - connections and relationships might break rather soon.

We probably should find ways to collapse and delete notifications that have become obsolete which is usually the most tedious part to find out after a longer absence...

Thursday, December 3, 2009

Doktorandenseminar ETH Zurich/Univ. St. Gallen

I just spent two full days at our group's (Elgar Fleisch) Doktorandenseminar. Each Ph.D. student had to present his current state of research. After three talks per session we did feedback walk-arounds such that each participant had the chance to comment - quite thrilling. I presented (slides) my understanding of research and tried to provide some lessons about publishing.
The event was very inspiring - despite spending time in the same office space we're still not of aware of all the projects going on...

Definitely a highlight were the lessons learned from Mikko Lehtonen (he just defended recently), he recommended to seek for a practical problem (motivation), start with an overview paper, circulate between own concepts and related work, design EU project reports as papers, define your stake in projects, write down your conclusions, copy from overview paper, copy from your papers - and the thesis is done! He obviously enjoyed Jakob Badram's Fish model presented at Ubicomp 2007's doctoral consortium.

There is no better way of getting a condensed view about what's going on in the brains of our twenty-something Ph.D.s and getting their views about my research in return!