cazmockett.com

Semantic Camp Session 6 – Parsing Microformats
Gareth Rushgrove

Gareth gave us a quick rundown of the various parsers which are available.

Language Specific – most available apart from Java.
Language agnostic – web services

hKit – open source PHP stuff.
Mofo – for Ruby
Sumo Javascript (from Dan Webb) – generic parser for JS.
XSLT – at Brian Suda’s site
Optimus – is down at the moment
Google’s Social Graph API – parses XFN relationships

Semantic Camp Session 7 – WTF is RDF?
Tom Morris

Always worth hearing, Tom did an idiots’ guide to RDF, just the thing for a bear of very little brain such as myself.

RDF = Resource Description Framework! 1999 originally, 2004 updated into 6 docs:

• Primer
• Schema
• Syntax Spec
• Concepts & Abstract Syntax
• Semantics
• Test Cases

RDF files at their simplest form:

3 things – subject, predicate, object (see above). The simplest form is – Triples (N3)
Subject and Predicate are Resources – Not literals; Object – Literal sting of text
foo.n3 – text file containing triples.

Tools to parse are available in – Java, C, Perl, Python, PHP, Ruby

He then talked some more about RDF validation and FOAF – which seems like the most friendly and usable aspect of RDF from this bear’s perspective. If you want to make your own FOAF file, why not use the foaf-o-matic

The Rest of the Afternoon

There were other talks going on, but I got distracted by a rather fab game, made by John Linklater-Johnson called Semantopoly. A fab idea [not in the least based on Monopoly] which had us all amused for some hours:

[Matt, Gareth and Isabelle get to grips with the rules of Semantopoly]

One of the presentations that I missed at BarCampLondon2 (I was attending another session) was a light-hearted debate about the similarities and differences between Microformats and RDF. The main protagonists were:

• Jeremy Keith and Brian Suda in the Microformats corner
• Tom Hughes-Croucher, Tom Morris and Ian Forrester in the RDF corner.

Thankfully, for those who didn’t see the debate, Ian has uploaded a video of the session. It makes interesting viewing! And shortly afterwards, I found Ben Ward’s insightful post about the whole subject too. I think Ben’s second paragraph hits the nail on the head:

The thing about RDF is that no-one has yet demonstrated any real-world reason to care about it. It fascinates academics who would love — just for the sake of it — to model the entire universe in triples but in the real world of web browsers the value has never really been promoted.

Spot on.

The Microformats advocates have been very quick to explain what they are for, what they do, and how to implement them. I use them regularly in this blog, and try to incorporate them wherever I can into new projects. It’s so easy to build them in from scratch when marking up events (hCalendar), people (XFN) or contact details (hCard).

But as yet, I’m really stumped as to what RDF – or more importantly, eRDF can do for me. Tom Morris has started a website called GetSemantic which hopes to chart the progress of developments about eRDF and spread the word. I’ll be keeping an eye on it from time to time, to see what’s cooking, but until then, I’ll be sticking to my diet of Microformats.

Introductions
We met in the plush surroundings of BT’s offices in St. Paul’s. They have some fantastic facilities and it was great to be able to make use of them for the weekend.

[hi-tech ramp down to the main auditorium]

The Three Musketeers – Ian, Nat and Jason – set the ball rolling with a few words and generally explained what was going on, and how the format of the weekend would run. Then we were all encouraged to say a few brief words about ourselves as an icebreaker, by passing the mic round the auditorium. I thought this was a great idea as you could note down someone’s name if you had similar interests, and seek them out later.

[the fun begins with the Three Museketeers introducing themselves]

Ian Forrester on Pipelines
Ian was explaining what Pipelines were about, basically inputting a stream of data (can be rss but also other formats), transforming via XSLT to give an output such as HTML. Several online applications currently let you do these sorts of mashups, eg BlogWaves 1.0 is a GUI application with which you can do transforms. Also, pipes.yahoo.com does a similar thing. Ian was disappointed that none of them seem to allow you to include information from your own desktop in the aggregation, which might also be useful.

http://www.touchstonelive.com – is a “glorified rss reader” plus extras. Allows you to set alerts as popup for desktop, or as an rss feed (via pebbles). Preloadr passes pictures through Flickr and out to Moo. All have APIs and so the pipeline can be automated right through, end to end, without the user having to intervene manually. Here’s Ian strutting his stuff:

[Ian talks about pipelines in the main auditorium]

Tom Morris on The Semantic Web
Tom’s talk was a tad too technical for a bear of very little brain like me. He started by saying that tagging data isn’t a very scaleable thing to do. RDF is a more complex way of tagging stuff. It can have the same design model as relational databases. And gives us a way to represent something like the relationship: Pants = Trousers – users of both terms are now able to understand each other. [It's helped now I've looked up the definition of ontology!]

eRDF is embeddable – ids, titles, attributes etc. It can be parsed with XSLT if your page is XHTML. RDFa is the next thing coming along with XHTML2.0

GRDDL (griddle), Protégé, SparQL are tools to help you do this stuff easily. SparQL lets you bypass APIs which would otherwise require all methods to have been published as an API.

[Tom talks triples]

After all that brain-bending stuff, it was lunchtime! Thank goodness.