Social media network bots identification and analysis

Dissertant

Abdi, Hamzah Jalal

Thesis advisor

al-Qaddumi, Ashraf Ahmad

Comitee Members

al-Bluwi, Ibrahim
Abu Tair, Diya al-Din I.
al-Atum, Jalal

University

Princess Sumaya University for Technology

Faculty

King Hussein Faculty for Computing Sciences

Department

Department of Computer Sciences

University Country

Jordan

Degree

Master

Degree Date

2015

English Abstract

With the rapid growth of Online Social Networks (OSN), their direct involvement in our daily lives and with the enormous number of users, OSNs have become the perfect target for people to use in a harmful way.

The OSNs are injected with huge number of fake accounts or automated accounts (bots) that try to spread malicious links and harmful data or try to benefit others on their agenda. In this thesis, a framework for Twitter is proposed.

Our proposed framework uses specific features extracted from Twitter profiles along with machine learning algorithms in order to identify bot accounts on Twitter.

The extracted features can differ between human’s behavior on Twitter and bot’s.

After extracting the necessary features, the machine learning algorithms produce a set of rules that can distinguish bot accounts from real accounts. Several tests has been conducted on a dataset collected by multiple methodologies for number of Twitter accounts.

The framework shows an amazing performance by having an accuracy of 96% and above in detecting bot accounts.

Main Subjects

Information Technology and Computer Science

Topics

• Social networking Web sites
• Security measures

No. of Pages

82

Table of Contents

• APA
• MLA
• AMA

Language

English

Data Type

Arab Theses

Record ID

BIM-650897