Wednesday Jan 13, 2010

Faviki: social bookmarking for 2010

faviki logo

Faviki is simply put the next generation social bookmarking service. "A bookmarking service? You must be kidding?!" I can hear you say in worried exasperation. "How can one innovate in that space?" Not only is it possible to innovate here, let me explain why I moved all my bookmarks from delicious over to faviki.

Like delicious, digg, twitter and others... Faviki uses crowd sourcing to allow one to share interesting web pages one has found, stay up to date on a specific topic of interest, and keep one's bookmarks synchronized across computers. So there is nothing new at that level. If you know del.icio.us, you won't be disoriented.

What is new is that instead of this being one crowd sourced application, it is in fact two. It builds on wikipedia to help you tag your content intelligently with concepts taken from dbpedia. Instead of tagging with strings the meaning of which you only understand at that time, you can have tags that make sense, backed by a real evolving encyclopedia. Sounds simple? Don't be deceived: there is a huge potential in this.

Let us start with the basics: What is tagging for? It is here to help us find information again, to categorize our resources into groups so that we can find them again in the rapidly increasing information space. I now have close to ten years of bookmarks saved away. As a result I can no longer remember what strings I used previously to tag certain categories of resources. Was it "hadopi", "paranoia", "social web", "socialweb", "web", "security", "politics", "zensursula", "bigbrother", "1984", ... If I tag a document about a city should I tag it "Munich", "München", "capital", "Bavaria", "Germany", "town", "agglomeration", "urbanism", "living", ...? As time passed I found it necessary to add more and more tags to my bookmarks, hoping that I would be able to find a resource again in the future by accidentally choosing one of those tags. But clearly that is not the solution. Any of those tags could furthermore be used very differently by other people on delicious. Crowd sourcing only partially works, because there is no clear understanding on what is meant by a tag, and there is no space to discuss that. Is "bank" the bank of a river, or the bank you put money in? Wikipedia has a disambiguation page for this, which took some time to put together. No such mechanism exists on delicious.

Faviki neatly solves this problem by using the work done by another crowd sourced application, and allowing you to tag your entries with concepts taken from there. Before you tag a page, Faviki finds some possible dbpedia concepts that could fit the content of the page to tag. When you then choose the tags, the definition from wikipedia is made visible so that you can choose which meaning of the tag you want to use. Finally when you tag, you don't tag with a string, but with a URI: the DBPedia URI for that concept. Now you can always go back and check the detailed meaning of your tags.

But that is just the beginning of the neatness of this system. Imagine you tag a page with http://dbpedia.org/resource/Munich (the user does not see this URL of course!). Then by using the growing linked data cloud Faviki or other services will be able to start doing some very interesting inferencing on this data. So since the above resource is known to be a town, a capital, to be in Germany which is in Europe, to have more than half a million inhabitants, to be along a certain river, that contains certain museums, to have different names in a number of other languages, to be related in certain ways to certain famous people (such as the current Pope)... it will be possible to improve the service to allow you to search for things in a much more generic way: you could search by asking Faviki for resources that were tagged with some European Town and the concept Art. If you are searching for "München" Faviki will be able to enlarge the search to Munich, since they will be known to be tags for the same city...

I will leave it as an exercise to the reader to think about other interesting ways to use this structured information to make finding resources easier. Here is an image of the state of the linked data cloud 6 months ago to stimulate your thinking :-)

.

But think about it the other way now. Not only are you helping your future self find information bookmarked semantically - let's use the term now - you are also making that information clearly available to wikipedia editors in the future. Consider for example the article "Lateralization of Brain Function" on wikipedia. The Faviki page on that subject is going to be a really interesting place to look to find good articles on the subject appearing on the web. So with Faviki you don't have to work directly on wikipedia to participate. You just need to tag your resources carefully!

Finally I am particularly pleased by Faviki, because it is exactly the service I described on this blog 3 years ago in my post Search, Tagging and Wikis, at the time when the folksonomy meme was in full swing, threatening according to it's fiercest proponents to put the semantic web enterprise into the dustbin of history.

Try out Faviki, and see who makes more sense.

Some further links:

Tuesday Jan 05, 2010

MISC 2010 and the Internet of Subjects

The International Conference on Mobility, Individualisation, Socialisation and Connectivity (MISC 2010) will be taking place in London from Jan 20 to 23 under the rallying cry "Personal Data It's Ours!". It will cover a very large number of topics in the space of Identity, the Social Web, Privacy and Data Ownership, (see the Agenda). I will be presenting on the developments of the Secure Social Web with foaf+ssl.

The conference will also be the launch pad for the Internet of Subjects foundation, whose manifesto starts with the following lines (full version)

The place digital technologies have now dwelled in our lives is leading to an ever-increasing flow of personal data circulating over the Internet. The current difficulties experienced in personal data management, like trust and privacy, are the revealing symptoms of a growing contradiction between an architecture that was primarily designed to manage documents, with the growing expectations of individuals of a more person-centric web. This contradiction will not be resolved by adding a simple patch to the current architecture; a second order change, similar to Copernican revolution, is required to move from a document-centric to a p erson-centric Internet, and create the conditions for a more balanced and mature relationship between individuals and organisations.

I completely sympathise with the feeling expressed by this message. But just as the Copernican revolution did not require an actual change in the movement of the planets - they have been turning around the Sun quite happily for billions of years - but 'only' required a change in how the humanity thought about the movement of the planets, so Web architecture as it currently stands, is perfectly adequate for an Internet of Subjects. It has been designed like that right from the beginning. Tim Berners Lee in his 1994 Plenary at the First International World Wide Web Conference, presented a Paper "W3 future directions" where he showed how from the flat world of documents as shown here

one could move to a world of objects described by those documents as shown here

This is what led to the development of the semantic web, and to technologies such as foaf that since 2000 have allowed us to build distributed Social Networks, and foaf+ssl that are allowing us now to secure them. Using the semantic web then to describe the authors of the documents and hence turn the web of objects into a web of subjects making statements about objects, does not require much technological innovation: it's built into the semweb architecture.

Still to someone who does not know this - the conference as well as the Manifesto are aimed at people who don't - their feeling will be that something is fundamentally wrong with web architecture. This is indeed the feeling the pre Copernican astronomers would have had as their models became more and more complicated to accommodate the always increasing amount of information they gathered about the stars. What should have been simple and beautiful, revealing the mind of God, must have seemed more and more confusing. Until one day, the way the world looked, suddenly changed...

Sunday Sep 20, 2009

Social Web Bar Camp in Paris

social web bar camp program drawn up on the black board

After flying in from Berlin on Friday and celebrating the Jewish new year late into the night with Ori Pekelman, I woke up earlyish on Saturday to go to the Social Web Bar Camp organized in and by La Cantine, the very friendly Parisian conference, community, meeting space for creative people in the digital age.

At 10am the conference started and people slowly arrived for the freely available espresso coffee and pastries. The conference was free too, being sponsored by the member organizations of La Cantine. At 10:20am as the coffee had worked itself into the 60 or more attendees, Ori started the workshop (picture) by having everybody introduce themselves shortly by name and 3 tags. The Bar Camp rules of the game were then explained:

  • Everybody is a participant
  • You make the event
  • Feel free to move between sessions if you feel you are not getting what you were looking for at one of them
  • Write up your interests on the black board, this will be used to create the time table.
So the sessions were put together on the spot there and then.

Of course I put up a session on foaf+ssl and Distributed Social Networks on the black board, for the session starting at 11am.

After a last coffee, a little over 20 people gathered in the room. I connected the laptop to the projector, introduced myself and the W3C Social Web XG, before starting the presentation (slides in pdf) which I have been giving in various universities and hacker spaces around Europe for the past 5 months. (see the FrOSCon video for example)

picture of the discussion in the foaf+ssl session

A round table discussion of this size has a very different dynamic to conference presentations. It is a lot more free flowing and people can ask question and did as I went through the presentation, leading to lively discussions on security, identity and web architecture. At times it seemed in danger of veering off into widely philosophical discussions, but somehow we always got back to the topic helped by the real implementations of foaf+ssl that are now available. Somehow we did in fact manage to complete covering the subject by 12:30 including an excursion into a description of the very real business opportunities this enables.

From the twitter posts (tagged #swcp) and the invitations to follow up with other French public and private institutions that I got over the course of the day, I can only say that this conference was a great success. I could not have started my 1 month stay in Paris in a better way. I will clearly be very busy during the coming month, before my return to Berlin.

Thanks to Huges M for the photos. More of his pictures are available on his flickr account under the #swcp tag.

Further pointers

About

bblfish

Search

Archives
« April 2014
MonTueWedThuFriSatSun
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
    
       
Today