Phishing Website Detection Using Machine Learning Python Project — Complete Documentation & Simulation

Overview & System Context

Phishing Website Detection Using Machine Learning Python Project: Comprehensive Web-Sourced Technical Investigation Exhaustive Research Dossier · Pure Web-Scraped Content: ~860 Words · Real Software Screenshots: 3 1. Executive Overview & Technical Background This project aims to detect phishing websites using machine learning techniques. The goal is to build a model that identifies phishing websites based on significant URL features and develop a user interface for real-time legitimacy checking.

Current systems train models based solely on the URL and cannot accommodate all website names, requiring enhancements in accuracy. Additionally, existing systems lack a specific user interface for easy access and usage. We developed a machine learning model that identifies phishing websites based on significant URL features and created a web interface for users to input URLs and check their legitimacy. If a URL is identified as a phishing website, the interface will display its legitimate counterpart and store the URL for further analysis.

I built an advanced model to detect phishing websites using a dataset of over 11,000 websites, each labeled as phishing or legitimate. By utilizing ensemble techniques like Random Forest and Gradient Boosting, I achieved exceptional accuracies of 97.4% and 97.9%, respectively. Figure 1: GitHub - sanjana4283/phishing-detection 2. System Architecture & Core Working Principles A phishing website is a common social engineering method that mimics trustful uniform resource locators (URLs) and webpages. The objective of this project is to train machine learning models and deep neural nets on the dataset created to predict phishing websites. Both phishing and benign URLs of websites are gathered to form a dataset and from them required URL and website content-based features are extracted. The performance level of each model is measures and compared.

The set of phishing URLs are collected from opensource service called PhishTank . This service provide a set of phishing URLs in multiple formats like csv, json etc.

that gets updated hourly.

To download the data: https://www.phishtank.com/developer_info.php . From this dataset, 5000 random phishing URLs are collected to train the ML models.

The legitimate URLs are obatined from the open datasets of the University of New Brunswick, https://www.unb.ca/cic/datasets/url-2016.html . This dataset has a collection of benign, spam, phishing, malware & defacement URLs. Out of all these types, the benign url dataset is considered for this project. From this dataset, 5000 random legitimate URLs are collected to train the ML models.

The above mentioned datasets are uploaded to the ' DataFiles ' folder of this repository. Figure 2: Detection of Phishing Websites Using Machine Learning | Python Final ... 3. Software Implementation, Frameworks & Protocols The details pertaining to these features are mentioned in the URL Feature Extraction.ipynb.

So, all together 17 features are extracted from the 10,000 URL dataset and are stored in ' 5.urldata.csv ' file in the DataFiles folder. The features are referenced from the https://archive.ics.uci.edu/ml/datasets/Phishing+Websites . Before stating the ML model training, the data is split into 80-20 i.e., 8000 training samples & 2000 testing samples. From the dataset, it is clear that this is a supervised machine learning task. There are two major types of supervised machine learning problems, called classification and regression.

This data set comes under classification problem, as the input URL is classified as phishing (1) or legitimate (0). The supervised machine learning models (classification) considered to train the dataset in this project are: Figure 3: Detecting Phishing Websites With Machine Learning | PDF | Phishing ... 4. Practical Deployment, Analysis & Case Studies All these models are trained on the dataset and evaluation of the model is done with the test dataset. The elaborate details of the models & its training are mentioned in Phishing Website Detection_Models & Training.ipynb The short video presentaion for this project is @ https://youtu.be/I1refTZp-pg . The slide presentaion used in this video is Phishing Website Detection by Machine Learning Techniques Presentation.pdf From the obtained results of the above models, XGBoost Classifier has highest model performance of 86.4%. So the model is saved to the file ' XGBoostClassifier.pickle.dat ' This project can be further extended to creation of browser extention or developed a GUI which takes the URL and predicts it's nature i.e., legitimate of phishing. As of now, I am working towards the creation of browser extention for this project. And may even try the GUI option also. The further developments will be updated at the earliest.

5. Performance Evaluation & Extended Technical Observations Phishing attacks involve deceptive websites and emails that mimic legitimate entities to steal sensitive information from users. These attacks are among the most common cybersecurity threats faced today. Machine learning offers powerful tools to automatically detect and flag these threats by learning from patterns in data. In this project, I apply three different machine learning models to a dataset of websites, aiming to classify them as either phishing or legitimate.

The dataset used in this analysis comprises various features of websites, including URL length, the use of HTTPS, the presence of an IP address in the URL, etc. It contains 10,000 entries, each labeled as '0' (legitimate) or '1' (phishing). This data was sourced from Kaggle , which compiled it to facilitate machine learning applications in cybersecurity. I evaluated three machine learning models: Logistic Regression, Random Forest, and Support Vector Machine. The goal was to determine which model provides the highest accuracy using parameter tuning and cross-validation techniques. The findings are summarized at the end of this project.

The Random Forest model outperformed others with an accuracy of approximately 96.8%, which implies that it could be the most suitable model for this particular dataset, likely due to its ability to handle nonlinear relationships and interactions between features effectively. Logistic Regression and SVM show competitive performance, but both are slightly lower than Random Forest.

Harvested Online Documentation & Reference Sources • https://github.com/pavanl1122/PHISHING-WEBSITE-DETECTION-USING-MACHINE-LEARNING • https://github.com/shreyagopal/Phishing-Website-Detection-by-Machine-Learning-Techniques • https://ishomarsnr.github.io/Projects/phishing_detection.html