Cybersecurity Project

Phishing Website Detection

A machine learning classifier that flags phishing URLs from structural, domain and content features before a user submits credentials.

Intermediate6–8 weeksCSE · IT · MCACybersecurityAI & Machine Learning

Project overview

The system extracts measurable features from a URL and the page behind it — length, subdomain depth, use of an IP address instead of a hostname, certificate age, form action targets, external resource ratio — and classifies the site as legitimate or phishing with a confidence value.

It is one of the best security projects for a student team: no illegal activity is involved, the datasets are public, and the feature engineering gives you plenty to discuss in a viva.

Problem statement

Phishing pages are cloned in minutes and taken down slowly, so blocklists are always behind. Users cannot judge a URL by looking at it, especially on a phone where the address bar truncates the domain.

Proposed solution

Build a feature extraction pipeline over URL structure, domain registration data and page content, train several classifiers, compare them on the same held-out split, and deploy the strongest as an API with a browser extension or web checker front end. Optimise the threshold for a low false positive rate, because blocking legitimate sites is the costlier error here.

Key features

  • Thirty-plus engineered features across URL, domain and content
  • WHOIS and certificate age lookups where available
  • Comparison of logistic regression, random forest and gradient boosting
  • Threshold tuning driven by false positive cost
  • Explanation of which features triggered a phishing verdict
  • Bulk URL scanning from an uploaded list
  • Optional browser extension that checks pages as they load

Technology stack

Feature extraction

Pythonurllib / tldextractpython-whoisBeautifulSoup

Modelling

scikit-learnXGBoostFeature importance analysis

Serving

Flask or FastAPISQLite

Client

ReactChrome extension (Manifest V3)

System architecture

The flow below is what goes into your architecture diagram and the corresponding chapter of the report.

  1. URL intake receives a link from the web checker or extension
  2. Feature extractor computes lexical features immediately and network features where reachable
  3. Model service loads the trained pipeline and returns a probability
  4. Decision layer applies the tuned threshold and produces a verdict
  5. Explanation layer lists the features that most influenced the score
  6. Result is cached so repeat lookups of the same domain are instant

Expected output

  • Phishing verdict with probability and contributing features
  • Model comparison table with precision, recall and false positive rate
  • Bulk scan report for a submitted URL list
  • ROC curve and threshold analysis for the report

Suitable branches, difficulty and duration

Suitable branches

Difficulty level

Intermediate. Comfortable for a team where at least one member has worked in this stack before.

Development duration

6–8 weeks including documentation, tested and ready to demonstrate. Milestones are planned around your review dates.

Documentation, review, demo and viva support

Documentation support

Abstract, problem statement, literature survey guidance, system architecture, ER and UML diagrams, module descriptions, test cases and screenshots — prepared in your department format.

Project review support

Delivery is split across your review calendar so review one has a design and plan, review two has working modules, and the final review has the integrated, tested system.

Demo support

A rehearsed demo path, seeded sample data, a reset script, and a recorded walkthrough kept as a backup in case something fails in the review room.

Viva preparation

A written question bank for this domain, a code walkthrough session, and practice at explaining each module in your own words — because the examiner will ask you.

Customisation options

Colleges reject duplicate topics, so most projects are adjusted before we start. These are the usual directions for this one.

  • Add screenshot-based visual similarity detection against known brands
  • Add email header analysis to cover phishing mail as well as pages
  • Add a reporting workflow so flagged sites can be reviewed and relabelled
  • Retrain on a feed of recently reported URLs to study concept drift
Talk to CodeLab69

Enquire about the Phishing Website Detection project

Send your branch, college and first review date. You will get confirmation of scope, what the customisation would look like, and a delivery plan mapped to your reviews.