Train SWAN with new training set

[bettinaballin]

Hi,

I'd really like to assess your approach with another dataset and some custom classes, as I'm currently doing some research on automatic SRM detection.

The documentation on your page states that in order to create the training dataset, the source JARs need to be processed with the two doclets (doc-coverage-doclet and doc-xml-exporter-doclet). Unfortunately, I couldn't get that part to work, as no comments/documentation is available. Which method do I need to run, which values are expected? The Main.java in doc-coverage-doclet shows this error: "Fehler: Keine Quelldateien für Package org.owasp.encoder".

Additionally, I cannot find the code you mentioned for Natural Language Processing in this project.

Many thanks for your help!