Indian Journal of Public Health Research & Development
Year : 2018, Volume : 9, Issue : 10
First page : ( 991) Last page : ( 995)
Print ISSN : 0976-0245. Online ISSN : 0976-5506.
Article DOI : 10.5958/0976-5506.2018.01299.8

Increase phishing tweets detection accuracy with improved classification features

Wooi Liew Seow¹, Sani Nor Fazlida Mohd¹, Abdullah Mohd. Taufik¹, Yaakob Razali¹, Sharum Mohd Yunus¹

¹Faculty of Computer Science and Information Technology, Universiti Putra Malaysia

Online published on 1 November, 2018.

Abstract

Phishing used to be well-known as a form of Social Engineering crime that deceives victims by directing them to a fake website where their confidential and sensitive information are gathered for unlawful activities. Traditionally, Phishing attacks targeted email users but have now exposed to target users at Online Social Networks (OSN)s such as Twitter, Facebook, Myspace, etc. Since OSNs have become so prevalent to Phishers to spread Phishing attacks, a research study to increase the Phishing tweets detection accuracy by improving the classification features tested on a dataset containing Phishing and safe Uniform Resource Locators (URL)s collected from Twitter is required. In this study, machine learning of Random Forest (RF) claimed to achieve high detection accuracy by other researchers and number of different classification features were used to test on a dataset collected from Twitter. The results of our experiment showed that we managed to identify 9 best classification features with higher achieved detection accuracy of 94.69% than 94.56% achieved by other researchers who using more than 9 features tested on the same dataset collected from Twitter.

Top

Keywords

Phishing, twitter, random forest, classification features.

Top