Article’s

Phishing schemes and junk emails represent a serious risk to internet users by deceiving them into disclosing confidential information. Current detection systems rely on blacklists and fixed rule-based filtering which often fail to detect newly developed malicious websites and misleading email content. This paper presents a machine learning-based intelligent system that can identify spam in emails and URLs. Preprocessing and TF-IDF vectorization are used in the process to extract features from supplied text and URLs. The inputs are classified as authentic or fraudulent using algorithms such as Random Forest and Logistic Regression. The goal of a Streamlit interface is to give users a simple interesting and easy-to-use experience. According to experimental results the recommended approach is more effective than traditional detection systems at revealing hidden phishing patterns. The developed model provides fast accurate and reliable forecasts for cybersecurity applications in real time.