Method for detecting fake news through writing style

Abstract

In the era of digital technologies, distinguishing truth from misinformation is a challenging task. Fake news, characterized by deceitful narratives, poses a significant threat. Traditional fact-checking methods often overlook the nuances of linguistic stylistic coloring. This study employs an advanced writing style analysis that extends beyond conventional methodologies. Several linguistic dimensions of texts are considered in this research, emphasizing on pre-processing and function development. The experiments are based on various datasets. Thus, the developed method for detecting fake news utilizes a multidimensional approach. The proposed development includes meticulous verification of the dataset, pre-processing, and function development, focusing on emotionally charged vocabulary, word groups used in reports indicating event likelihood, mild cursing, and non-standard lexicon. Significant differences in linguistic features were identified, contributing to a nuanced understanding of the construction and creation of deceptive texts. The research results demonstrate that this method accurately distinguishes genuine from fake news articles based on writing style. This study represents significant progress in identifying phony news through writing style analysis, aiding in combating misinformation in the era of digital technologies.