[foaf-protocols] A personal introduction to this mailing list

Benjamin Heitmann benjamin.heitmann at deri.org
Wed Sep 8 17:35:32 CEST 2010


I would like to introduce myself to this mailing list, and introduce my current main interested regarding WebID. 

My name is Benjamin Heitmann, and I am a PhD student at DERI Galway. 
My current research interests include software architecture, adaptive personalisation and the evolution of the Web of Data.

I would like to contribute towards the specification of the parts of WebID which are concerned with authorising access to private resources. 

Why am I interested in this? I have a specific use case in mind, however explaining it might take a few minutes of your time ;) 

Below is a short textual description, however I also gave a presentation about this specific topic, and the slides are at [1].

Amazon, Last.fm and Facebook all have recommender systems, which try to show you new and interesting stuff. 
However over the years the role of recommendation has changed: 
* Previously a recommender system only tried to sell you items from the inventory of the same website you are on.
* Today recommender systems will try to find content (music, news) or advertisements for you which are not limited to the stuff that is available from one site only. 

You all probably know about the Facebook Open Graph. What is the business motivation behind that? 
Facebook has created the Open Graph architecture in order to pull as much data about the user into his profile as possible, 
in order to make the recommendations for the user as relevant as possible. 
Then it can match advertisements to users in the same way that it recommends other things to the user.

This development points to three (or four) paradigm shifts: 
0.) Recommendations have become a commodity, and are required for certain types of services. 
1.) recommendations will come from beyond the context of just one site
2.) recommendations will be outside of only one domain (not just books or music, but any relevant topic)
3.) privacy is an important factor in gaining the trust of the user, so that the user gives you data to make recommendations in the first place.

The research problem: 
In order to get profile data from the user while at the same time maintaining the privacy of the user,
somehow the privacy of the user profile data needs to be maintained in a decentralised way. 

If you look at this from the user perspective, then the user wants a portable profile that is based on open standards
while maintaining his privacy at the same time. 

And thats where WebIDs come into the discussion :) 

WebIDs can be an enabling technology for creating open ecosystems in which users can share their data with different services, 
while at the same time controlling which service can access which parts of their profile. 

And that is the use case which motivates my interest in standardising the authorisation of private resource through the use of WebIDs. 

I will introduce some ideas on the authorisation issue in a second email. 

Thanks to everybody who was reading this far *g

cheers, Benjamin. 

[1] http://www.slideshare.net/metaman/what-your-hairstyle-says-about-your-political-preferences-and-why-you-should-care-about-the-future-of-recommender-systems

More information about the foaf-protocols mailing list