Fake News Detection Using Machine Learning And Knowledge Graph

In today’s digital era,fake news presents serious problems.It rapidly spreads over social media,plays on confirmation bias,and undermines the credibility of reputable news publishers.Machine learning(ML)refers to the systematic exploration of algorithms and statistical models,while deep learning enables the design of computational models that consist of numerous layers of processing that can acquire abstract representations of data at various levels.1)This study focuses on different approaches to machine learning and compares different models to evaluate and analyze their accuracy.Subsequently,a hybrid model is proposed,featuring a nine-layer architecture to achieve accurate binary sequence classification.The use of various layers allows for effective feature extraction,dependency capture,and attention focusing,while dropout regularization prevents over-fitting.The proposed model has achieved an accuracy of 97.4%on a large dataset,which outperforms existing models like LSTM and CNN.2)Secondly,a knowledge graph is constructed to visualize fake news detection results.The knowledge graph categorizes news items as true or false using labels 0 and 1,respectively.Instead of creating separate nodes,our knowledge graph employs a linking approach for enabling the examination of dissemination patterns and the identification of publishers which publish the two types of news.3)Finally,a recommendation system is developed to recommend the news detection results to the readers.The proposed hybrid model is employed within the recommendation system.If the news is true,the recommendation system automatically recommends the reader to go to the website for further reading;otherwise,if the news is false,it does not encourage the reader to read it.