Archive for March, 2008

RecSys 2008

Monday, March 31st, 2008

The CFP for RecSys 2008 is out: http://www.recsys.acm.org/. Paper submission is May 18th, 2008.

The doctoral symposium ad has also been circulated: http://recsys.acm.org/doctoral.html

The Recommender Systems 2008 Doctoral Symposium provides an opportunity for doctoral students to explore and develop their research interests in an interdisciplinary workshop, under the guidance of a panel of distinguished research faculty. We invite students who feel they would benefit from this kind of feedback on their dissertation work to apply for this unique opportunity to share their work with students in a similar situation as well as senior researchers in the field. The strongest candidates will be those who have an idea and an area, and have made some progress, but who are not so far along that they can no longer make changes. Typically, this means they will have made their dissertation proposal, but still be about a year from completion.

Trustlet

Friday, March 28th, 2008

When browsing around the blogs I read, I came across trustlet: a wiki site dedicated to sharing scientific research on trust metrics in social networks. It includes an excellent list of conferences/workshops that deal with trust, and an incredible list of links to datasets – ranging from wikipedia, email networks, and blog networks, to the much sought advogato and epinions datasets. Much more than we have done on our own dataset page!

What a gold mine!

SAC TRECK 2008

Friday, March 21st, 2008

I’m glad to say that the TRECK track of SAC went quite well and did not suffer from some of the things I mentioned in my previous rant. The track was organized by Dr. Jean-Marc Seigneur of the University of Geneva, and the two sessions were chaired by Dr. Virgilio Almeida of the Federal University of Minas Gerais (who I had an interesting discussion with after the track), and was broadly divided into two themes: trust and recommender systems. The trust session had an overall focus on peer-to-peer systems, here are some quick samples:

  • Francesco Santini presented the idea of multitrust, which aims at computing trust in a dynamically created group of trustees who all have different subjective trust values ["Propagating Multitrust Within Trust Networks, " Bistarelli/Santini].
  • Asmaa Adnane presented the application of trust to detecting misbehaviour in link-state routing algorithms. I always wonder how well these cool ideas will work in practice; if information is lost or delayed they will deduce that another node is untrustworthy! ["Autonomic Trust Reasoning Enables Misbehavior Detection in OLSR," Adnane/Timoteo de Sousa/Bidan/Me']
  • The Surework Framework extended the current operation of trust in p2p networks to include the idea of super-peers; nodes with very high reputation can, in fact, become reputation servers. ["Surework: A Super-peer Reputation Framework for p2p Networks," Rodriguez-Perez/Esparza/Munoz]
  • The CAT Model was introduced and explained- it is a model of open and dynamic systems that considers services as contexts.. The 15 minute time-limit was a bit constraining and I’ll have to read the full paper!  ["CAT: A Context-Aware Trust Model for Open and Dynamic Systems" Uddin/Zulkernine/Ahamed]
  • Rowan Martin-Hughes applied a game-theoretic analysis to understand why people would defect in a large-scale open system, like eBay. The analysis was based on a modified version of the Prisoner’s dilemma, which was very interesting; the only question that arises is, as Daniele mentioned, is this appropriate when users may very well behave irrationally? ["Examining the Motivations of Defection in Large-Scale Open Systems," Martin-Hughes/Renz]

The second session focused on recommender systems:

  • Karen Tso-Sutter presented her work on combining user-item tags into the collaborative filtering process. Interestingly, tags did not improve accuracy until the algorithm was already boosted by using both user- and item- based algorithms. ["Tag-Aware Recommender Systems by Fusion of Collaborative Filtering Algorithms," Tso-Sutter/Marinho/Schmidt-Thieme]
  • My work! Looking at the similarity distribution over a graph generated by a nearest-neighbour algorithm. ["The Effect of Correlation Coefficients on Communities of Recommenders," Lathia/Hailes/Capra].
  • Patricia Victor‘s paper discussed an extension to Paolo Massa’s work on trust-aware recommender systems, which concluded that the cold-start problem in recommender systems can be avoided by having users express trust values in other users, which can then be propagated. The problem is: which users should they connect to? The paper has an interesting analysis of the different kind of users in the epinions dataset. ["Whom Should I Trust? The Impact of Key Figures on Cold-Start Recommendations," Victor/Cornelis/Teredesai/De Cock].
  • The last paper veered away from collaborative filtering to look at the role of keywords and taxonomies in content-based recommender systems. The taxonomy vs. folksonomy war continues! ["Comparing Keywords and Taxonomies in the Representation of Users Profiles in a Content-Based Recommender System" Loh/Lorenzi/Simoes/Wives/Oliveira]

The full list of abstracts can be read on the trustcomp-treck web site. If any of the attendees or authors are reading this post: we welcome your thoughts and comments, and officially invite you to contribute to this blog! To write a guest-post about your research, please get in touch! (n.lathia @ cs.ucl.ac.uk)

Algorithm finds the network – for genes or the Internet

Thursday, March 20th, 2008

Maybe someone is interested:

“In a recent paper in Physical Review E 77:016104 (2008), Weixiong Zhang, Ph.D., Washington University associate professor of computer science and engineering and of genetics, along with his Ph.D. student, Jianhua Ruan, published an algorithm to automatically identify communities and their subtle structures in various networks.”
(news article).

Franco

No-Shows and PowerPoint Karaoke @ ACM SAC 2008

Wednesday, March 19th, 2008

I’m currently at ACM SAC 2008, in beautiful Fortaleza (Brazil). The conference has a number of tracks; many of them are very interesting and they cover a very broad range of topics. I’ve jumped around through a few different sessions, including CISIA (Computational Logic and Computational Intelligence in Signal and Image Analysis), WT (Web Technologies), IAR (Information Access & Retrieval), SWA (The Semantic Web and Applications), and MMV (Multimedia and Visualisation). The TRECK track, where I am presenting, is on the last day of the conference.

Looking at the titles of the papers to be presented, each session promised to be very exciting; and even if I am no expert in the field I wanted to see what I could get out of it. Unfortunately, in a lot of the sessions I have attended as many as 3/5 of the presenters did not show up, leaving the session chairs in a very embarassing situation. In one case, a session ended early and so I went to the next room. I got there in time to hear the chair ending that session as well due to no-shows: very dissapointing. I’ve seen presenters leaving a session once their bit is done, and not contribute to discussions that they are supposed to be experts in, leaving only the session chairs to ask questions.

To make matters worse, some of the presentations I have seen have been quite poor. They all follow the same formula: title, outline, a bunch of slides where the presenter reads what is on the slide, conclusion where they repeat what they just presented, and future work: power point karaoke at its best. I understand that some may follow this method due to difficulties with English, but they do not even sound interested in their own work. They also immediately dive into the details or cite other papers without explaining why, forgetting that some (like me) may be there to learn something new and need the broad strokes of the picture first. (I’m no expert at presenting; however, there are lots of resources on presenting on Daniele’s page here). It makes attendance at the sessions even more difficult!

Overall, the conference has been organized well, and is running smoothly. The track chairs obviously put a lot of work into this, and it should be an ideal opportunity to mingle with a range of researchers… but the participants are not doing a very good job! I’ll write a separate post about the interesting presentations I have seen.

Turn Ideas into Money

Monday, March 17th, 2008

As the GroupLens research blog is reporting, MyStrands have announced a $100,000 investment for the winner of the recommender startup competition. The winners will be announced at RecSys 2008. On a side note, any UCL-ers interested in entrepreneurship might also be interested in this course run by the UCL graduate school.

Update: it seems that tapping into any wisdom hidden in the masses is the new source of ideas (crowdsourcing): don’t come up with ideas, just make a means for the ideas to come to you. A new competition is adding its name to the Netflix prize, this previous post on evaluating algorithms with the masses, and the above competition: semantihacker is offerring $1 million to anyone who can put their semantic-analysis engine to good use!

Researching new mobile applications

Saturday, March 15th, 2008

“With the new iPhone SDK, it’s just a matter of time before we see a wave of new Web 2.0 applications.” Here are 12 Future Apps For Your iPhone (which may well inspire our  research agenda):

1. Reality Tagging
2. People Tagging
3. Reality Recognition
4. Physical Social Networks
5. Personalized Travel Guides
6. Digital and Physical Treasure Hunt
7. Distributed Mobile Games
8. Credit Card and Biometrics as Software
9. Paperless Receipts & Digital Business Cards
10. Medical records as Software
11. Physical Browsing & Digital Shopping
12. Location/time-based deals

Evaluating our smart algorithms

Saturday, March 15th, 2008

Many of us are designing smart algorithms and are often supposed to evaluate them by carrying out well-designed user studies.

Problem: Those studies are expensive and, consequently, we tend to trade off between sample size, time requirements, and monetary costs.

Proposal (by PARC researchers) : To collect user measurements from micro-task markets (such as Amazon’s Mechanical Turk). Here is their blog post (which comments on their upcoming HCI paper titled “Crowdsourcing User Studies With Mechanical Turk“).

Note: Often, users are irrational.

Forever Connected (for free)

Saturday, March 15th, 2008

There is an online map of all the free wifi in London. Maybe one day it will be free everywhere? Of course that flares the debate of net neutrality: who is going to maintain the companies that maintain all the Internet infrastructure? What will be the consequence of these companies charging for their services- will the fusion of media and medium lead to a downturn of the web? There seem to be a flare of hot opinions going around: what do you think?

Complex Data Course

Thursday, March 13th, 2008

This week the UCL Graduate School ran a series of lectures on modelling complex data- something that any researcher will encounter, but may often not be prepared to do or understand the tools that are being used. These lectures were run by some really smart people from UCL Gatsby and the Centre for Computational Statistics and Machine Learning, and there are a number of online resources including notes, slides, example data sets and matlab code:

  • Regression (lost the link!)
  • Classification/Supervised Learning (page)
  • Dimensionality Reduction techniques (page)
  • Clustering (zip file)

Think and (Don’t) Speak

Thursday, March 13th, 2008

Following up on recent posts about ubiquitous computing, here is a video that I saw: a neckband that can process the signals your brain sends to your vocal chords and can speak for you! The video is not the best quality, but is interesting nonetheless.. (no more listening to someone else shouting down their mobile phone?)

http://www.youtube.com/watch?v=xyN4ViZ21N0

From computers to ubiquitous computing, by 2020

Tuesday, March 11th, 2008

On the 17th (Tuesday) and 18th of March, a research event titled “From computers to ubiquitous computing” will take place at The Royal Society in London. Many world-class speakers. Abstract: The convergence of wireless networks with new sensing technologies and devices has started to embed computers into our everyday life. This meeting will consider this shift to ‘Ubiquitous Computing’ as an interdisciplinary grand challenge affecting all aspects of computer science that has massive implications for how we might reason about, build and experience computer systems with considerable issues for society. Register here.

Mobiles+Smart Wine Racks= Collaborative Wining!

Saturday, March 8th, 2008

This beatiful “smart wine rack uses radio frequency identification (RFID) to track individual bottles in the rack and identifies ones that fit the users’ wine selection criteria. … A handheld device accepts queries, and full-color LED lights transform the elegant wine rack and the bottles themselves into a search results interface.” (here)

Cool! What’s next? Collaborative wining, of course. :-) Phones store wine preferences and smart wine racks run collaborative filtering algorithms for recommending wines ;-)

P2P Lending (good web 2.0)

Saturday, March 8th, 2008

Also known as Person-to-person lending. Borrowers and lenders come together directly on the web and do not need banks. Thanks to Mo, I recently found out that that P2P lending is “working its way into the charitable sector: Kiva.org puts potential “social investors” in touch with small businesses in the developing world, which promise to send e-mail updates on how the business is developing.” Here is how it works (see this video presentation for more): (more…)

Tagging on the move

Saturday, March 8th, 2008

Soon I will team up with Licia and Valentina to study how mobile users may retrieve digital content using tags. We will do so by combining folksonomy and reputation(trust) systems. I will then present our preliminary results at the session on Trust in Percom of Secrypt. For more, pls keep an eye on the comments below this post that will appear at the end of July. In the meantime, here is a fine piece of research on how to automatically tag what people see through their cyber googles ;-)