Notes from the Wandora workshop:

On the tutorial day of the Topic Maps 2009 conference I attended the Wandora workshop — lead by Olli Lyytinen of Grip Studios Interactive, Finland.

It was cool to see – and try – how easy it is to create mash-ups of information using some of the many Wandora extractors, connecting bits and pieces of information to a unified view.

As part of the workshop, we created a (tiny) Monty Python ontology filled with information from sources such as YouTube, Wikipedia, IMDB, BBC’s RSS, etc.

Wandora uses layers of virtual topic maps to separate different information that is merged when viewing it. That way one can for instance separate base information and information extracted from the external sources like Wikipedia, YouTube, Flickr, IMDB, RSS feeds, multimedia metadata, etc.

What was new to me, who’d barely looked at Wandora before (limited to download + open and have a look), is that Wandora can also be used as a Topic Maps engine (for use in e.g. web applications). In fact, Wandora also has a built-in server which can be used to browse Wandora topic maps.

The interface of the Wandora editor is a bit cluttered and developer centric / difficult to use, although it seems to have a lot of features. After extracting data from a source, you can for instance “transform” information items to new topics, etc. As an example, Olli first used a directory extractor to get MP3 files from a folder, then he selected some of the created File instances and chose to extract new information using these as subject locators for an MP3 ID3 extractor. That way, you can for example easily create a topic map of your MP3 collection.

Combine this with information from last.fm, YouTube, etc., and voila … you can connect the dots.