Automated Fake News Detection -1.pptx
- 1. Automated Fake News Detection with supervised learning approaches and NLP processing techniques Batch Members Guided By: Arunkumar A (813018104006) Ms.G Sathy Vincent A (813018104044) AP, CSE Yugesh R (8130181040 Oxford Engg College
- 2. ABSTRACT Fake news detection attracts many researchers attention due to the negative impacts on the society. We propose a novel fake news Logistic regression technique.
- 3. INTRODUCTION In order to distinguish fake news ,we investigate the one of largest fake news dataset which called LIAR dataset. To enhance the performance on fake news detection task we use Logistic regression.
- 4. DOMAIN MACHINE LEARNING It is a type of AI that allows software application to become more accurate at predicting outcomes.
- 5. LITERATURE SURVEY Title Author, year, Journal Method or algorithm Advantage Disadvantage Supervised Learning for Fake News Detection Julio C. S. Reis et al, 2019, IEEE XGBoost (XGB) AUC = 0.86 Need to increase the accuracy Beyond news contents: The role of social context for fake news detection Kai Shu et al, 2019, WSDM tri-relationship embedding framework has been used Better overall performance Need to increase the accuracy of the model
- 6. Fake News Detection on Social Media using Geometric Deep Learning Federico Monti et al, 2019, IEEE geometric deep learning 92.7% ROC AUC More system complexity Multi-Source Multi-Class Fake News Detection Hamid Karimi et al, 2019, Multi_x0002_source Multi-class Fake news Detection framework has been proposed Easy to implement The accuracy of the model is 38.81
- 7. EXISTING SYSTEM In the existing method, fake news detection multi-task learning (FDML) model has been used. FDML model investigates the impact of topic labels for the fake news and introduce contextual information of news.
- 8. DISADVANTAGE Low accuracy. Need to increase the over all performance of the model. The existing model unable to detect the fake news for different dataset.
- 9. PROPOSED SYSTEM In this method, we propose Fake news detection technique with logistic regression architecture. For the pre processing, NLP processes are perform to extract the information from the text data. The proposed architecture is deployed in web based application by Django framework.
- 10. ALGORITHM LOGISTIC REGRESSION Logistic regression is one of the most popular Machine Learning algorithms, which comes under the Supervised Learning technique. It is used to calculate or predict the probability of binary event occuring
- 11. PROPOSED SYSTEM In this method, we propose Fake news detection technique with logistic regression architecture. For the pre processing, NLP processes are perform to extract the information from the text data. The proposed architecture is deployed in web based application by Django framework.
- 12. ARCHITECTURE DIAGRAM Dataset Data collection NLP processing Logistic regression Model saving Dataset Model deployment Heroku
- 13. MODULE LIST Data collection NLP Processing Model selection Model Deployment
- 14. DATA COLLECTION Data for our project is collected form the website called Kaggle. Kaggle is a open source repository for data scientist and machine learning enthusiasts.
- 15. NLP Processing The text data not usable to build the machine learning model. Hence,We need to convert into the vector format. In this process, the NLP (Natural language processing) toolkit is used to perform the vectorization operations.
- 16. Model Selection In the proposed method, the Logistic regression is used to build the model. The based model is saved to deploy the model.
- 17. Model Deployment Flask is a frame work of the python to build the website. The flask is used to deploy our model with user friendly. Heroku is a cloud platform which is used to host our website for the online users.
- 18. CONCLUSION The proposed method Logistic regression based system presented higher accuracy while compared with other existing approaches. The accuracy of the model is 97.5% on the test data. The proposed model has ability to perform the classification operation on different datasets.
- 19. ADVANTAGE Higher accuracy of the model. The performance of the model is high. The proposed model has ability to work with different kind of dataset.
- 20. REFERENCES Reis, J.C., Correia, A., Murai, F., Veloso, A. and Benevenuto, F., 2019. Supervised learning for fake news detection. IEEE Intelligent Systems, 34(2), pp.76-81. Shu, K., Wang, S. and Liu, H., 2019, January. Beyond news contents: The role of social context for fake news detection. In Proceedings of the twelfth ACM international conference on web search and data mining (pp. 312-320). Monti, F., Frasca, F., Eynard, D., Mannion, D. and Bronstein, M.M., 2019. Fake news detection on social media using geometric deep learning. arXiv preprint arXiv:1902.06673. sKarimi, H., Roy, P., Saba-Sadiya, S. and Tang, J., 2018, August. Multi- source multi-class fake news detection. In Proceedings of the 27th international conference on computational linguistics (pp. 1546-1557).