TY - JOUR ID - TI - Machine Learning-Based Effective Detection Scheme of Fake News AU - Khalid Abood Kamel , Jumana Waleed PY - 2022 VL - 14 IS - 3 SP - 155 EP - 163 JO - Journal of Al-Qadisiyah for Computer Science and Mathematics مجلة القادسية لعلوم الحاسوب والرياضيات SN - 20740204 25213504 AB - Today, extremely large amounts of false news are consistently uploaded by malevolent people with fraudulent intentions, endangering democracy, justice, and public confidence while having highly harmful social impacts on both individuals and society. Due to the inherent uncontrollable posting processes of social media sites (such as Facebook, Twitter, and Snapchat), this is particularly pertinent to them. The need for earlier false news identification has considerably fueled efforts in academia and business to create more accurate fake news detection techniques. Unfortunately, there isn't much information available regarding how news spreads. There are benefits and drawbacks to relying on social media to follow the news. Social media platforms do make it possible for information to flow swiftly among users. However, these websites could be used to disseminate "fake news," which is low-quality content that contains errors. The widespread dissemination of false information has an extremely damaging effect on both people and society. As a result, the detection of fake news posted on numerous social media platforms has recently emerged as a highly regarded field of study. In this paper, an effective detection scheme of fake news based on the commonly utilized techniques of machine learning has been presented. This proposed scheme involves diverse phases; Dataset preprocessing phase, extracting the phase of the feature, and Naive Bayes (NB) and K-nearest Neighbor are used in the categorization step (KNN). The obtained results in this presented scheme exhibit that the utilization of the NB classification technique exceeds the K-NN classification technique with an accuracy of 94% using the same dataset.

ER -