Copyright © 2025 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{165791,
        author = {Ravi Prakash and Shambu Kumar Singh},
        title = {False News Detection from Text on Social Media Using K-Nearest Neighbors Bayesian Approach},
        journal = {International Journal of Innovative Research in Technology},
        year = {},
        volume = {11},
        number = {1},
        pages = {1611-1620},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=165791},
        abstract = {The pervasive spread of false news on social media presents significant threats to public discourse and societal stability. This study investigates advanced methodologies for the detection of false news specifically within social media contexts, leveraging the unique characteristics and challenges posed by these platforms. A diverse dataset is compiled from various social media sources, including platforms like Twitter, Facebook, and Reddit, consisting of both verified true news and identified false news. The preprocessing pipeline is tailored to handle the noisy and informal nature of social media text, including techniques such as tokenization, slang normalization, hashtag processing, and the handling of emojis and special characters. Feature extraction methods, including term frequency-inverse document frequency (TF-IDF), word embeddings, and advanced contextual embeddings (e.g., BERT, RoBERTa), are employed to capture the linguistic features of social media text. Additionally, network-based features, such as user interactions, repost patterns, and user credibility scores, are integrated to enrich the feature set. The proposed detection framework leverages the K-Nearest Neighbors algorithm to identify patterns and similarities in the feature space, while a Bayesian approach is integrated to provide probabilistic assessments of news veracity. This hybrid method aims to combine the strengths of KNN in capturing local data structures with the probabilistic reasoning capabilities of Bayesian models. The performance of the proposed method is evaluated using standard metrics such as accuracy, precision, recall, and F1-score. Comparative analysis with traditional machine learning models and deep learning approaches is conducted to validate the effectiveness of the KNN-Bayesian hybrid model. Experimental results demonstrate that the KNN-Bayesian approach achieves competitive performance, with significant improvements in detection accuracy and robustness compared to baseline models. The incorporation of metadata and contextual information further enhances the model's ability to discern false news from authentic content. Challenges such as the dynamic nature of false news, the need for scalable real-time detection, an},
        keywords = {False News/Information Detection, K-Nearest Neighbours, Bayesian, Word2Vector, Term Frequency- Inverse Document Frequency.},
        month = {},
        }