Anti-Spoofing in Medical Employee's Email using Machine Learning Uclassify Algorithm

Since the advent of COVID-19, healthcare and IT cybersecurity have been an issue. Digital services foreign labor increased cyberattacks. July 2021 saw 260,642 phishing emails. 94% 12 countries’ employees experienced epidemic Phishing attacks steal sensitive data from spam emails or legitimate websites for profit. uses URL, domain, page, content variables. Simple machine-learning methods stop This study discusses patient employee accounts cybersecurity. paper covers COVID-19 email detection. article examines message's subject, email, links. Uclassify classifies content, spam, languages automates Semi-supervised machine learning dominates healthcare. The algorithm used multinomial Naive Bayesian classifiers. Document class is [0–1]. compared Multinomial in two experiments with other algorithms. Experiment 1 achieved MNB accuracy 96% based on a database Kaggle Phishing. 2 showed that system accurately predicted URL hyperlink targets PhishTank data. 96.67% respondents correctly identified URLs, 91.6% did so hyperlinks. These focused Tokenization, Lemmatization, Feature Extraction (FE) contained internal feature set (IFS) external (EFS). more exact than earlier since it decimal digits word frequency. only takes binary inputs. can detect spoofing.