dkpro / dkpro-similarity

Word and text similarity measures
https://dkpro.github.io/dkpro-similarity
Other
53 stars 22 forks source link

getting started with dkpro-similarity is not really satisfying #48

Closed helt closed 8 years ago

helt commented 8 years ago

The gettings started guide on https://dkpro.github.io/dkpro-similarity/gettingstarted/ does not really gets up to the point, where it is clear to the reader (i.e. to me), how the code actually looks like, which is in the magic rectangle (the rect with the bold border...)

(My guess is that i iterate over another Collection within an AnalysisEngine, but havent found much yet, i.e. in a getNext(JCas cas) i have a loop which fetches other JCas objects)

Any hints would be appreciated.

text_sim_uima_pipe

zesch commented 8 years ago

It depends a bit on what you are trying to achieve with DKPro Similarity. Some code examples are here: https://dkpro.github.io/dkpro-similarity/

The getting started guide might be misnamed a bit, as it tries to show how to reproduce our SemEval 2013 results.

For examples how to use the different measures, I recommend a look in the unit tests.