Submission File Layout

Your software must generate a file (.csv) similar to Submission File Layout (please check the table in the left side). The file should have a table with two columns and 400 rows. Since there are 400 samples in the test dataset, therefore, number of rows are 400. The first column in the table corresponds to the file number and the second column is the classifier output (Real/Fake). Please note that every file in the test dataset should be classified.

News Article file number, Label

Please check the table (Submission File Layout ) as a sample.

For example, the first row in Submission File Layout (please check the table in the left side):

1 Corresponds to file number input news article, and F corresponds to the predicted label of the news article.

Submission notes

1. In order to encourage the investigation of different kinds of features, a team can submit only 3 different runs as the team want. However, the best run results will be considered as the final run.

2. You can submit your runs by sending the generated files to UrduFake2020 (at) CICLing .org with your full name.

3. Once the results are available, participants will be required to provide an abstract and a technical report including a brief description of their approach and experiments for the publication in the FIRE Proceedings. Technical papers should be of a maximum of 4 pages (excluding references) following CEUR Proceedings and will be published in CEUR Workshop Proceedings.