Skip to content

Latest commit

 

History

History
15 lines (9 loc) · 791 Bytes

README.md

File metadata and controls

15 lines (9 loc) · 791 Bytes

Urdu-Fake-News

A machine translated corpus of news articles translated from English to Urdu news articles. This corpus is not yet publically available for research purpose. It is submitted to some journal for publication. After publication, the corpus would be made publically availble for research purpose.

Corpus Properties are given below: Number of Real Articles = 1032

Number of Fake Articles = 968

Maximum Article Lenght = 7045

Minimum Article Lenght = 25

Vocabulary Size = 2,101,801

The models of Machine learning (SVM, NB, and DT) and ensemble learning models (Ensemble Selection, Cascade Generalization, Voting, Grading, and Stacking ) on both corpora (UFN and BET) are given in individual folders. All the models are generated using open-source data mining software WEKA.