But I have to ask, why are we still using this approach, which has known inaccuracies and quirks (in English)? Its one virtue is compactness, which in the old days was more important than accuracy.
I think porter2 works well enough. You don't need lexical category like lemmatisation, so it's a lot simpler than that approach.You also don't need a huge volume oftraining data.
You don't need any training data to do better than Porter2, just lexical data. If you're content to map the same surface form to the same stem regardless of context, as Porter2 would, what you need is a dictionary mapping words to their stems.
This used to be considered a lot of memory to use. It's not anymore.
This leaves you with some edge cases (is "axes" the plural of "axis" or "axe"?), but it's much better than mixing up the words "universe" and "university".
For those interested in the Stemming subject in Elixir, I made a simple Elixir wrapper around snowball implementations of stemmers : https://github.com/awetzel/stemex
9 comments
[ 8.8 ms ] story [ 206 ms ] thread[1] https://github.com/fredwu/simple_bayes
But I have to ask, why are we still using this approach, which has known inaccuracies and quirks (in English)? Its one virtue is compactness, which in the old days was more important than accuracy.
But it is fun, I'll grant you that.
This used to be considered a lot of memory to use. It's not anymore.
This leaves you with some edge cases (is "axes" the plural of "axis" or "axe"?), but it's much better than mixing up the words "universe" and "university".
snowball -> C -> NIF -> Elixir