SmiDCA: Smishing attack detection for mobile computing on smishing dataset

Nowadays nearly everyone is using mobile computer/devices such as smart-phones and laptops to conduct their business transactions and for social purposes. While this trend has significantly transformed working and personal lifestyles worldwide, it has also led to serious concerns about threats to security and privacy among individuals as well as organizations. One of the most widespread security threats is phishing attacks launched for the purpose of stealing certain sensitive information of victims and then abusing this information to illegally obtain confidential data. There are many types of phishing attack such as social phishing, spear-phishing, pharming, and smishing. Recently Joo et al. (2017) proposed an improved security prototype to detecting Smishing attack on mobile computing known as S-Detector. Their model is able to distinguish between normal SMS message and phishing. However Goel and Jain (2017a) claimed that S-Detector does not address three SMS security message features. First, S-Detector cannot not check for login page within the SMS message. Second, it is not efficient in detecting self-answering messages and Lastly, text normalization is not achieved. To solve these issues (Sonowal and Kuppusamy, 2018) propose new technique called SmiDCA. In this research, we re-implement SmiDCA using dataset called smishing dataset for Harm ans Spam (Almeida, 2017). The re-implement SmiDCA technique is analyzed SMS messages and extracted the security features of SMS to detect the smishing SMS messages efficiently.

Text FSKTM 2019 15 - IR.pdf Download (1MB)

Actions (login required)

View Item