               Tennant: Digital Libraries
Recent Posts
Recent Comments
Most Commented On
Archives
Blog
Link This | Email this | Blog This | Comments (5)
The Semantic Web, Part II: Linked DataJanuary 7, 2009 Yesterday I attacked what I believe to be some somewhat unbelievable predictions for the Semantic Web. But today I set out to shine a spotlight on a part of the Semantic Web that I think holds real promise. It's called "linked data".One of the reasons I like "linked data" is because the name pretty much says it all, and instead of calling to mind intelligent robots circling the Internet making decisions for you it seems to set out a pretty simple goal -- to allow data to be linked together in useful ways. Now that's a practical goal I can get behind. If it so happens that eventually when we get enough data linked up some of the things the Semantic Web folks talk about may become possible, then fine. But the point is we don't need to hold our breath waiting for it, we can set our goals at more practical levels and get some real work done today. Linked data is more exposed on the web than data made available via APIs -- crawlers can crawl it, applications can link to it, it can link to other open data, etc. It is much more discoverable and usable to applications that may not know how to access a specific API. There are a growing number of data sets that are being made available this way, including some library-specific sets of data. For example, Ed Summers at the Library of Congress recently experimented with putting the Library of Congress Subject Headings up as linked data. He brought it down recently in anticipation of the official exposure of this data by LC (possibly to be released as soon as this month), but it provided an interesting window into how to do something like this and how it could be made available in a useful way on the web. Before he took it down I captured some screenshots so you can see what it was like. Keep in mind that you are viewing the human-readable versions of data that is actually meant to be accessed and used by software -- but today's web technology makes it relatively easy to expose the data in both machine and human usable ways simultaneously. Another example comes to us courtesy of Martin Malmsten of the National Library of Sweden. This year he made the Swedish Union Catalog, LIBRIS, available as linked data. It may not be obvious from these data sets exactly how a growing universe of linked data can enable useful new connections between related data sources, but I am increasingly convinced that making data available in this fashion can only enable interesting (and likely unforseen) uses of this data. I still believe we are very far away from the kind of future envisioned by Semantic Web cheerleaders, but the work of people in the trenches doing useful things like Ed Summers and Martin Malmsten (among many others) with linked data gives me hope that there will be some good things to come. I'll be watching developments closely, as I suggest many of you do as well (or even better, participate). If you're interested, there are several places where you can get started:
Posted by Roy Tennant on January 7, 2009 | Comments (5) Industries: News & Features
January 8, 2009
In response to: The Semantic Web, Part II: Linked Data Erik Hetzner commented: Hi Roy -
January 8, 2009
In response to: The Semantic Web, Part II: Linked Data Jeff Young commented: I think that the principles of Linked Data need to be liberated from their RDF assumptions, and I've started to blog about the possibilities on q6.oclc.org. In particular, I believe that UML class diagrams are a powerful alternative to the traditional graph model. Fortunately, the models are not mutually exclusive.
January 10, 2009
In response to: The Semantic Web, Part II: Linked Data Michael Hausenblas commented: Roy, thanks a lot for this. I think you rightly captured the linked data gist and nicely described its value. Whereas we focused on *building* datasets in 2007 and 2008, now the community slowly starts to contemplate on how to exploit these datasets, that is building application on top of it. Please have a look at [1] (currently mainly a stub, but we work on it ;) and share your thoughts there as well!
January 13, 2009
In response to: The Semantic Web, Part II: Linked Data Callie Bowdish commented: It would be nice to continue on with this conversation with were Web Ontology Language fits into all of this.
January 15, 2009
In response to: The Semantic Web, Part II: Linked Data Antti Seppänen commented: Check out finnish CultureSampo: www.kulttuurisampo.fi/?lang=en
Advertisement
|
Advertisements
|
|
|
|
