The start of the semantic web?
The Xpragmatic View #100
April 21, 2008
by Marc Buyens, Xpragma
Download as PDF-file 
A couple of weeks ago, Twine started in private beta. Over the past months, Twine has been hyped as being the first real application for the so-called semantic web, often also referred to as Web 3.0. Are we at the start of something big or is this just more of the same thing?
Since a couple of days, we started using Twine. Twine is one of the more hyped web applications of the past months and is viewed by many as the first real Web 3.0 or semantic web application.
It is a bit difficult to exactly describe what Twine is offering, since the solution covers quite a bit of different things:
- First, it is a social bookmarking service, where you can store and tag anything that you find on the Internet that is of interest to you, similar to a service such as Del.icio.us.
- In a similar way, you can upload or e-mail documents, images, videos and other data that again, you can tag and provide short descriptions for.
- While bookmarking or uploading content, Twine will parse this information automatically for the names of people, places, organisations, etc. adding additional tags where applicable.
- This information can be held private or shared, e.g. with people you have connected to. As the latter suggests, Twine is also a social networking platform, be it that the focus is not on the individuals, but on the content they favour.
- Collected items then can be linked into what we call a "context", or a "twine" in the jargon of the platform. Twines can be seen as categories, but here, a category does not "own" the item. Items can be linked to whatever number of contexts (or twines).
- When a twine is open for the public or a group, it essentially becomes a forum, allowing participants to comment, post additional items, etc.
- Based upon the items you collected and the twines where you participate, Twine will then recommend other twines or other members with matching interests. These recommendations are dynamic, so that when visiting a specific twine, the recommendations will specifically relate to that twine and to the members that match that context.
All in all, this provides for a very flexible mechanism to organise information, to discover new information that is of interest to you and to find like-minded people.
This said, being a beta product, not all is perfect. We had hoped that Twine would have parsed the complete document that you bookmark, but this is not (yet) the case. Other weaknesses of a beta product are of course also present and therefore, opinions are somewhat mixed, as you can read in this posting of ReadWriteWeb, titled Twine Disappoints After Semantic Web Hype.
Still, we are more inclined agreeing with the viewpoint of Hrafn Thórisson of Think Artificial in the article Why I Migrated Over to Twine (And Other Social Services Bit the Dust) which also provides some more detailed description of the product.
We also have already started migrating certain content to Twine. Still on a limited basis, but this after only using the service for a couple of days...
The reason for doing so is essentially the different "feeling".
One of the main reasons for the potential success of Twine and, today, also one of the main reasons for failure has to do with the fact that Twine invites you to abandon present habits. Twines might look like categories or forums, but they are not. They are perspectives on information.
These perspectives are not static either. Basically, a twine is a search for certain content. It is a representation of information in a specific context. Once you get that perspective, you can further drill down into new perspectives by using the tags that are linked to this content.
What you will find as the remaining context or perspective is not necessarily equal to what would have resulted from your own classification of content. Which representation of information is the more intelligent one?
That will be the real challenge for Twine. There are a lot of technical and usability shortcomings in the current product, but none of these are fatal. The real challenge will be educating users in making use of information in a different way.
How different? We don't know, but we started exploring...
Categories: Semantic web (Web 3.0)
RSS feed: http://www.xpragma.com/english/rss/xpven.xml
