Description of different extractors? about boilerpipe HOT 3 CLOSED

The CanolaExtractor has been trained on Canola documents. Its main purpose is 
to demonstrate the competitiveness of such a simple classifier (based on number 
of words/densities) for the Canola corpus evaluation. I would not recommend it 
for other purposes.

I'd recommend using ArticleExtractor for any type of news articles and 
DefaultExtractor (or maybe LargestContentExtractor) for the rest. YMMV.

I have provided some [Benchmarks 
http://code.google.com/p/boilerpipe/wiki/Benchmarks] on the L3S-GN1 news 
corpus, as an initial starting point.