Pakistani media fake news classification using machine learning classifiers

Irfan Kareem, Shahid Mahmood Awan

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

17 Citations (Scopus)


Identification of Fake News is import now a days because it is affecting our social life and opinions. Public misinformation detection is complicated task especially Pakistani media Fake News classification. We have seen Fake News in every aspect of life like politics, sports, business, entertainment and many more. For identification of fake news, we have done popular news websites scrap and develop our corpus of 344 News articles and labeled it manually Fake or True. We have investigated two feature extraction techniques like Term Frequency (TF) and Term Frequency-Inverse Document Frequency (TF-IDF). Seven different supervised Machine Learning (ML) classification algorithms are used and their results comparison have done. Best performance classifier K Nearest Neighbors (KNN) gives 70% accuracy and logistic regression gives 69% accuracy. Results can improved further by increasing number of articles in corpus.
Original languageEnglish
Title of host publication2019 International Conference on Innovative Computing (ICIC)
Place of PublicationPiscataway, NJ
ISBN (Print)9781728146829
Publication statusPublished - 23 Jan 2020


Dive into the research topics of 'Pakistani media fake news classification using machine learning classifiers'. Together they form a unique fingerprint.

Cite this