I am working on a process to parse Mediawiki markup from the german Wiktionary with the goal to automatically generate a (semantic) word net for nouns, similar to WordNet and ConceptNet. I am far from finished, but I was able to break down the markup and now I am going to extract the noun info.

The following images show a small tool I wrote to watch the processing steps. You can load a word by it's ID from the Wiktionary database (dumped and loaded into a MySQL table) and watch all data.