"An Efficient Approach For Detecting Phishing Web"

Project Abstract / Summary : Our Project is "An Efficient Approach for Detecting Phishing Web" which presents an automatic approach for bright phishing web detection based on learning from a large number of legitimate and phishing webs. As given a web, its Uniform Resource Locator (URL) features are first analyzed, and then classified by Naive Bayesian (NB) classifier. When the web’s legality is still suspicious, its webpage is parsed into a document object model (DOM) tree, and then classified by Support Vector Machine (SVM) classifier.
Phishing is the attempt to acquire sensitive information such as usernames, passwords, and credit card details (and sometimes, indirectly money) by covering as an authentic entity in electronic communication. Communications declare to be from popular social web sites, auction sites, banks, online payment processors or IT administrators are commonly used to lure unsuspecting public. A typical phishing attacks consists of four phases. Namely preparation, mass broadcast, mature and account hijack. To detect and prevent various kinds of phishing attacks, there are many different preventive strategies and detective ideas Information security specialists and anti-phishing organizations have set up phishing alerts databases that assess each reported phishing incident in terms of its risk level.

In order to detect phishing web, first NB classifier detects the URL features , they are stated as bellow:
URL Features:
Feature extraction plays an important role for the efficient prediction of phishing web.

Dots in URL: If Many dots are present in URL then that URL is phishing one.

Suspicious URL: If the sign like @ and - present in given URL then that URL is Suspicious.


Slash in URL: The URL should not contain more number of slashes. If it contains more than five slashes then the URL is considered to be a phishing URL.


After analyzing the URL features by using NB classifier the webpage is found to be still suspicious then, further SVM classifier is used to detect the webpage features in order to detect the phishing web.


Webpage features:

The feature vector generated in this step would then be inputted into a SVM classifier to determine whether a web is a phishing or a legitimate web.


Forms: Most phishing web contains forms asking for confidential information like username, password, credit card number etc. otherwise criminals are not getting personal information they want.


Nil anchors: It is an anchor which points nowhere.


Foreign Anchor: If domain name in URL is not similar to page URL then it is called as foreign anchor.


Foreign requests: It isSimilar to foreign anchor.



Foreign Anchor in Identity Set: It points to foreign domain.


Foreign request in Identity set: Here domain is compared with the URL identity.


SSL Certificate: It creates encrypted connection between web server and user’s web browser.


The dataset used for learning is collected from PHISHTANK. The dataset with 600 phishing webs and 400 legitimate webs is developed for implementation.100 legitimate and 100 phishing

Webs are taken as the training set, and the rest of 300 legitimate and 500 phishing pages compose the testing dataset.

Why did you choose to work on this project topic : Network-based security threats have led to widespread identity theft and financial fraud.In today's era, there is a threat to security in many online transaction these System will work together with the Classifiers based methods to provide better protection. It is a system which is use to check phishing website, it will not allow the sharing of personal confidential information such as password, credit card, username information and lot many. It will solve the problem of various online attacks. It will preserve the security.

Project Category : CS / IT / Networking
------------------------------------------------------
Institute/College Name: S.R.E.S. College of Engineering Kopargaon
City: Kopargaon
State: Maharashtra
Participating Team From: Final Year