Detection of False Information on Social Media

As of 2024, precise figures regarding the proportion of false information on Indian social media remain elusive. However, past data sheds light on the prevalence of misinformation. A January 2022 poll found that 22% of Indian social media users admitted to being deceived by fake news online. Another survey revealed that 45% encountered entirely fabricated stories in the Indian media, often with political or economic motives. Additionally, an Oxford University study noted that 54% of Indians relied on social media for truthful information. These figures, though estimations, underscore the significant presence of false information. Moreover, various definitions of “fake information” exist, encompassing dormant accounts to bot-driven spamming, highlighting the complexity of the issue. The dissemination of inaccurate information on social media platforms poses a substantial threat to online discourse integrity. This paper proposes a framework for detecting fake information utilizing common Python libraries like NumPy, Pandas, Matplotlib, NLTK, Joblib, and LDA alongside machine learning techniques. Preprocessing of textual data employs NLTK for natural language processing, followed by topic modeling with LDA to uncover latent themes. Machine learning algorithms integrated with NumPy and Pandas extract features and train models for post classification. Visualization tools like Matplotlib and Seaborn aid in data exploration and result assessment. This interdisciplinary approach demonstrates promising capabilities in identifying false information on social media platforms, contributing to ongoing efforts to combat online disinformation.