Effective Benign Identity Prediction Among Deceptive Accounts using Machine Learning

Everybody's life now revolves on virtual entertainment. Every day, the majority of people spend their time on Social Media Platform(SMPs). The number of profiles on social networking sites is rapidly growing, but users conceal their identities for malicious purposes. Contrarily, Machine learning techniques have been used on multiple occasions to correctly identify accounts that were created by bots. Detection of deceptive accounts is mandatory in the current situation because, in many fields, decision making is done through reviews generated by the account holder in SMPs. In order to obtain quality outcomes that are generated based on this kind of analysis, effective benign accounts should be predicted effectively. Reviews generated on the Twitter platform are gathered as a dataset, and classification is done based on tweeted content,including profile information.In our proposed method, Decision Trees, XGBoost Classifier,Logistic Regression,Random Forest Classifier and Support Vector Machine(SVM) algorithms are utilized.The research attempts to effectively identify false identities generated by people in Twitter using the same features previously used as well as grammar check is done for the tweeted content.