This repository contains all different types of data from the GermEval-2018 shared task.
The manually labelled GermEval-2018 data consists of two files:
- germeval2018.training.txt
This file represents the training data from the shared task. It contains the annotation for both tasks of the shared task.
- germeval2018.test.txt
This file represents the test data from the shared task. It contains the annotation for both tasks of the shared task.
Detailed guidelines regarding the annotation scheme of the data can be found in the following file:
guidelines-iggsa-shared.pdf
lIf you publish any work using the GermEval-2018 data, please cite the following publication:
Michael Wiegand, Melanie Siegel, and Josef Ruppenhofer: "Overview of the GermEval 2018 Shared Task on the Identification of Offensive Language", in Proceedings of the GermEval, 2018, Vienna, Austria.
The file
evaluationScriptGermeval2018.pl
represents the evaluation tool. It can be used to evaluated the predictions for both tasks of the shared task.
In order to obtain further information on the usage of that tool please type:
perl evaluationScriptGermeval2018.pl --help
The complete evaluation of all runs submitted by participants can be found in the file:
results.pdf
The submissions of all participants for both tasks can be found in the package:
submissions.tgz
All participants had to complete an online survey as part of their submission. The results of the survey are available in the file:
survey.pdf
More information about the shared task can be found on the website of the task: https://projects.fzai.h-da.de/iggsa/
Please direct any questions that you have about this data to the organizers of the shared task:
Josef Ruppenhofer email: ruppenhofer@ids-mannheim.de
Melanie Siegel email: melanie.siegel@h-da.de
Michael Wiegand email: michael.wiegand@lsv.uni-saarland.de