Monday, January 30, 2006

Weekly PhD Report 06, week 4

Work done:


- Presented the topic of my research interest at HMDB lunch


- Kept on reading research papers on different recommender systems and issues


- Finalised the chapter for the LOR review


Things to do:


- Continue by paper reviews


- Continue to work on my research questions and establish a plan for it


- Start preparing the work on a social information retrieval book chapter that was accepted in a call for contributions. See an abstract below


Towards Interoperable Social Recommendation: Storing, Sharing, and Reusing Evaluations of Learning Objects



Abstract



Since the early days of the use of digital learning resources in teaching and learning, quality has played an important role, sometimes more implicitly than explicitly. In the domain of digital learning resources, a plethora of ways to evaluate learning resources exists. However, limited ways to share and reuse these evaluations for social information retrieval purposes currently exist. Social recommendation is based on the opinions of the members of a community, in order to help individuals in that community more effectively identify content of interest from a potentially overwhelming set of choices.



This chapter emphasises the importance that the representation and storage of user experiences and evaluations of learning resources has, as a means to support social recommendation, and help users search the Web for learning resources in an effective and efficient manner. Thus, this chapter proposes a scheme for storing results from learning resources’ evaluation in a structured and reusable way, so that they can be shared between social recommendation systems. More specifically, the chapter develops an Evaluation Learning Object Metadata (evaLOM) model, a metadata schema that allows the description of the evaluation method used for learning resources, as well as storing its evaluation results. This model is part of a larger specification called Learning Object Log Specification (LO Log).

Friday, January 27, 2006

World according to World66

World66 ("the travel guide you write") is a great place, and it has become better and better over the years (still in perpetual beta, I see). There is this funny tool that let's you create a world map of countries that you've visited.

Mine looks like this:



Create your own visited country map
or check our Brussels travel guide

Tuesday, January 24, 2006

Weekly PhD Report 06, week 3

Work done:
  • I worked on my initial and preliminary research questions
  • Continued my paper review on different types and applications of recommender systems
  • Worked on my study plans
  • Worked a day in EUN and on my plan for 2005
  • Blogged about the new round on EU patent law: http://flosse.dicole.org

Things to do remains pretty much the same:
  • Continue by research reviews
  • Continue to work on my research questions
  • Give on more round on the LOR book chapter

Wednesday, January 18, 2006

Weekly PhD Report 06, week 2

Work done this week:
  • Participated in the HMDB lunch about clustering
  • Worked on the bookchapter about LORs.
    • The good news: the deadline has been extended until the end of the month, so we'll have time to adapt it according to some comments (Thanks Jehad and Martin!)
    • I'll send a new version around later this week in case someone wants to give it another try
  • Continued my paper review on different types and applications of recommender systems
  • Started working on my more precise research questions, hurray!
  • Checked a new application called http://www.yelp.com/ It's a social networking tool where you can find, review and talk about what's going on in your neighborhood, meaning in SF and a few other big cities in the US.
    • They have it all what you need nowadays for a YASNS (Yet Another Social Networking Service http://en.wikipedia.org/wiki/YASNS), you can add your network of friends, profiles and a Firefox toolbar for a direct search.
  • Tried to convince my promoter to do my pending administrational paper work

Things to do:
  • I'll start at EUN again, only one day a week and with less to do than last year (oh yeah!)
  • Continue my research reviews
  • Continue to work on my research questions
  • Continue to follow up the EC Patents discussion now with Open Consultation round
  • Put hard pressure on my promoter to finalise my pending administrational paper work.
  • Work on the LOR book chapter

Wednesday, January 11, 2006

Weekly PhD Report 06, week 1

work done:

  • Started an extensive review on recommendation systems, especially looking for papers on LORs and social networks
    • I'm looking especially design requirement and needs for LO recommender system, but interested in dilemmas found in the field and open questions such as what would the benefit of having a distributed service over a centralised one, etc.
  • Worked on the book chapter on LORs that I sent you a while ago.
  • Worked on a plan for the year and looked into some possible collaboration paths with people, namely picked up on an idea to write something with David Tosh from ELGG.
  • Tried to revitalise the plan for NING developments in vain
  • Kept Furling my reading and findings: http://www.furl.net/members/vuorikari

work to be done:

  • Continue my reading and note taking on recommendation systems
  • finalise the book chapter
  • write a scenario for Jehad on using the logs for LO review purposes
  • work on a plan to run a pilot on Ning
  • keep up the good spirit

Weekly PhD Report 51-52

  • Spent time writing a part for a paper "An Overview of Learning Object Repositories" with Argiris Tzikopoulos, Nikos Manouselis. Argis has made a survey on 59 repositories, and I was asked to help writing a part of it. I wish to send you this paper for comments later this week.
  • Worked on a proposal for a book chapter which would be called "Towards Interoperable Social Recommendation: Storing, Sharing, and Reusing Evaluations of Learning Objects". The idea would contribute to Jehad&all's Log specifications. This would also be used for recommendation purposes. See the attachment for more information.
  • Played around with swicki, see: http://riina1-swicki.eurekster.com/
    • Swickis are community search engines that learn from users

  • Checked out http://digg.com, the new hot tool for the community to review news. This is an interesting post about Digg and some other relevant social software.
    http://thomashawk.com/2006/01/digg-delicious-kaboodle-and-wink.html
    • The kick in Digg is that when you submit a link there, it will be Dugg [imperfect] by other news readers. As soon as one post starts getting Diggs on it, i.e. it is read by many other who think it's relevant, it will climb higher up on the relevance ranking list, thus appear on the front page of the Digg.
  • Worked on Ning.com application on social recommendation with Julien, Ben and Bram. We were not able to make it search Ariadne yet, but should be feasible soon.
    • The issues seems to be that we added our, well, Julien's own SOAP library to make the connection. However, the Amazon services module in Ning is pure PHP running in the Playground, as are the modules that use other web services (such as Flickr and Gmail). they said "You should be able to make external HTTP requests from Ning just fine, whether using curl or PEAR's HTTP::Client (both of which we support)." Hmmm...



Weekly PhD Report 51-52

  • Spent time writing a part for a paper "An Overview of Learning Object Repositories" with Argiris Tzikopoulos, Nikos Manouselis. Argis has made a survey on 59 repositories, and I was asked to help writing a part of it. I wish to send you this paper for comments later this week.
  • Worked on a proposal for a book chapter which would be called "Towards Interoperable Social Recommendation: Storing, Sharing, and Reusing Evaluations of Learning Objects". The idea would contribute to Jehad&all's Log specifications. This would also be used for recommendation purposes. See the attachment for more information.
  • Played around with swicki, see: http://riina1-swicki.eurekster.com/
    • Swickis are community search engines that learn from users

  • Checked out http://digg.com, the new hot tool for community reviewing of news.
  • Worked on Ning.com application on social recommendation with Julien, Ben and Bram. We were not able to make it search Ariadne yet, but should be feasible soon.