Copyright © 2026 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{207562,
        author = {Pooja Jalikatti and Prof. Shahina Indikar and Vaibhavi Renake},
        title = {Intelligent Detection of Misinformation in Digital Media Using NLP-Based Text Classi-fication and Machine Learning},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {13},
        number = {3},
        pages = {1611-1619},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=207562},
        abstract = {The rapid growth of digital communication platforms and online news portals has significantly increased the availability and sharing of information across the world. However, this advancement has also created a major challenge in identifying and controlling the spread of fake news. Fake news contains misleading or false information that can influence public opinion, create confusion, and reduce trust in digital media sources. Traditional manual verification methods are time-consuming and inefficient due to the large amount of news content generated every day. Therefore, an automated fake news detection system using Artificial Intelligence and Machine Learning techniques is re-quired to classify news information accurately and effi-ciently. This project proposes a Fake News Detection System using DistilBERT Transformer and Machine Learning Algorithms to identify whether a given news article or text content is fake or genuine. The system applies Natural Language Processing (NLP) techniques for understanding textual information and extracting meaningful features from news data. Initially, the input dataset containing news text and corresponding labels is collected and processed. The text preprocessing phase removes unwanted information such as URLs, special characters, symbols, and unnecessary content. The cleaned text is then converted into numerical feature representations using the DistilBERT-based Sentence Transformer model. Unlike traditional feature extrac-tion techniques such as Bag of Words and TF-IDF, transformer-based embeddings capture the semantic meaning and contextual relationship between words, which improves the overall prediction performance.
After feature extraction, multiple machine learning algorithms including Logistic Regression, Support Vec-tor Machine (SVM), and Random Forest Classifier are trained using the generated DistilBERT embeddings. The dataset is divided into training and testing sets to evaluate the performance of each model. The models are analyzed using different evaluation metrics such as accuracy, precision, recall, and F1-score. Based on the comparison results, the best-performing model is auto-matically selected using the highest F1-score and stored for future prediction. This approach ensures that the system uses the most reliable classifier for detecting fake news content. A user-friendly web application is developed using the Flask framework to provide real-time fake news prediction. Users can enter any news text through the web interface, where the system pre-processes the input, extracts transformer-based,features,and predicts whether the information is fake or genuine using the trained model. The applica-tion also provides a confidence score for the prediction result. Additionally, SQLite database integration is implemented to store previous prediction records, al-lowing users to view detection history. The proposed system improves fake news identification by combining advanced transformer-based language representation with efficient machine learning classification tech-niques. The integration of DistilBERT improves the ability of the system to understand complex text pat-terns and provides better classification compared to traditional approaches. This project demonstrates the effectiveness of artificial intelligence in reducing misin-formation and supports users in verifying the authen-ticity of digital news content. In the future, the system can be enhanced by integrating real-time news verifica-tion, multilingual fake news detection, and advanced deep learning models for improved accuracy and scala-bility.},
        keywords = {DistilBERT, NLP, Text Classification, Semantic Em-beddings, Misinformation Analysis, Machine Learning, News Verification.},
        month = {August},
        }