A Multistage Framework to Defend Against Phishing Attacks

A Multistage Framework to Defend Against Phishing Attacks

Madhusudhanan Chandrasekaran (SUNY at Buffalo, USA) and Shambhu Upadhyaya (State University of New York, USA)
DOI: 10.4018/978-1-60566-132-2.ch011
OnDemand PDF Download:
$30.00
List Price: $37.50

Abstract

Phishing scams pose a serious threat to end-users and commercial institutions alike. E-mail continues to be the favorite vehicle to perpetrate such scams, mainly due to its widespread use combined with the ability to easily spoof them. Several approaches, both generic and specialized, have been proposed to address this growing problem. However, phishing techniques, growing in ingenuity as well as sophistication, render these solutions weak. To overcome these limitations, we propose a multistage framework – the first stage aims at detecting phishing based on their semantic and structural properties, whereas in the second stage we propose a proactive technique based on a challenge-response technique to establish the authenticity of a Web site. Using live e-mail data, we demonstrate that our approach with these two stages is able to detect a wider range of phishing attacks than existing schemes. Also, our performance analysis study shows that the implementation overhead introduced by our tool is negligibly small.
Chapter Preview
Top

Introduction

Phishing is a form of Web based attack where attackers employ deceit and social engineering to defraud users of their private and confidential information such as password, credit card number, social security number (SSN), and bank account number. As the Internet is becoming the de facto medium for online banking and trade, phishing attacks are gaining notoriety, especially amongst hacker communities. Anonymity over the Internet, coupled with the potential for large financial gains serves as strong motivation for attackers to perpetrate such seemingly low risk, yet high return scams. The first recorded mention of phishing attacks was in AOL forums (“Phishing - Wikipedia,”) wherein attackers posing as system administrators tricked the registered users into disclosing their account information. Since then, phishing attacks growing in sophistication and ingenuity have affected millions of users causing heavy monetary damage. For example, in the year 2006 alone, phishing attacks cost $2.8 billion in losses to consumers and commercial organizations worldwide (Gartner Press Release, 2006).

Due to its widespread adoption and ability to be easily spoofed, email continues to be the favorite vehicle to perpetrate such scams. Email based phishing attacks are usually carried out as a three step process: (i) In the first step, phishers harvest email addresses of their potential victims from Web pages, online forums and by other social engineering mechanisms; (ii) For the second step, a large volume of specially crafted emails appearing to originate from legitimate domains is dispatched to the assimilated list using open SMTP servers and compromised machines. These emails contain hyperlinks which redirect the users to a fake Web site similar in appearance to the legitimate domain; (iii) Finally, account details and other personal information are collected from the users who unsuspectingly provide them into the fake Web site thinking it to be a legitimate one. Phishing attacks, like other social engineering attacks, for their success depend upon users’ lack of system knowledge. Phishers adopt a variety of visual deception agents to imitate the legitimate Web site’s look-and-feel (Drake, Oliver, & Koontz, 2004). The mimicry of a legitimate Web site is usually achieved through spoofing the URLs with non-ASCII Unicode characters using customized images to mask fake URLs and embedding the fake Web sites within images that resemble a browser window. Recent studies (Dhamija, Tygar, & Hearst, 2006) show that naïve users are inept in identifying common browser based cues such as address bar, status bar, SSL certificates, and toolbar indicators and often fall prey to such imitation sites.

Until recently, anti-spam techniques were employed to detect phishing emails. However as phishing emails closely resemble their legitimate counterpart, they do not share similar features as that of spam emails. Also, there exist a vast number of readily available tools that can bypass both the statistical and rule based spam filters. Several browser extensions and plug-ins have been proposed to detect phishing attacks. Although these techniques act as a first line of defense, they suffer from many limitations. First, as these approaches operate on the fake Web site, they take the users a step closer to the attack giving little leeway for suspicion. Second, most of the existing defense mechanisms are not automated and delegate the onus of decision making onto the users. Third, as these tools embrace the authenticity of the IP address as an important classification criterion, they fail to protect from attacks that are launched within the realm of legitimate domain. For example, an attacker could compromise a Web server and launch phishing pages from the domain itself1.

Key Terms in this Chapter

Context Models: Context models encapsulate the messages conveyed in the phishers’ emails to attract the potential victims into the fake Web sites. Phishers usually employ some kind of threat, fake reward, and false pretext in their email message to trick the users.

Phishing Email Structural Properties: Phishing email structural properties represent the set of invariant features that are present in most, if not all, of the phishing emails. These invariant properties are mostly visual deceptive agents employed by the phisher to trick the users. These invariant properties also helps in building discriminators that are accurate and less prone to false positives.

Challenge-Response Analysis: Challenge-response analysis is an authentication mechanism where either one or both the communicating parties adhere to a pre-agreed protocol used in verifying their identities. The party which desires to prove its identity has to provide correct response to the challenge posed by the opposite party with which it desires to communicate.

Email / Web site Spoofing: Email/Web site spoofing is the process by which the look-and-feel and the behavior of fake Web sites/emails is forged to mimic their legitimate counterpart.

Linear Binary Classification: The process of separating a set of m examples {(x1, y1)… (xm, ym)} into two regions by a linearly separable hyperplane parameterized by w such that yi (xi . w + b) > 0 for all i = 1…m. Such a hyperplane is called as separating hyperplane.

Feature Selection: Feature Selection is a process of selecting a subset of relevant features so that the net performance of underlying classifier is increased. Feature selection helps to minimize the presence of “noise” that adversely affects the model building.

Phishing: Phishing is a form of Web based identity theft where attackers employ deceit and social engineering to defraud users of their private and confidential information such as password, credit card number, social security number (SSN), and bank account number.

Complete Chapter List

Search this Book:
Reset