Comparative experiment
llooFlashooll opened this issue · comments
Hi, Professor. I saw your four repositories of Cujo
, Zozzle
, JaSt
and this JStap
. You really make a good implementation. I want to run some comparative experiments. Do these four repos have the same dataset files construction? JStap has a demo.pdf for launching. I'm wondering whether the others have the same dataset(Good/Bad, train/valid/test) setup to launch?
Through my test, set a valid dataset path then I can launch. Thanks!
Hi! Thank you for reaching out.
Could you solve your issue? You can definitely train all detectors on the datasets I shipped with JStap.
Yes! I succeed. Thanks for your timely reply!
But there's a tiny problem with the Rieck et al. with Cujo repo. I raise an issue on that repo: Aurore54F/lexical-jsdetector#1. Mainly the error with Jassi.
Sure no problem :)
Alright let's discuss that in the other repo then.