Automated Detecting Spammers In Social Media

Abstract

Social networks provide a way for users to remain in contact with their friends. The increasing popularity of social
networks allows social site users to gather large amounts of individual information about their friends. Among numerous sites,
Twitter is the fastest growing website. Its popularity has also attracted many spammers to use large amounts of spam to penetrate
legitimate users' accounts. In this research work, the Spam detection system in social sites” is designed to detect the spammer by
using a machine learning approach. Initially, data is collected from H-Spam14 site and then different pre- processing schemes
such as to convert data into lowercase; stop word removal will be applied. After this, the data enters into the feature extraction
phase, in which tokenization process is used to divide the entire sentence into a group of words and hence extract the best features
from the raw data. To select an appropriate value of extracted feature set, Artificial Bee Colony (ABC) has been applied as an
optimization algorithm to determine the optimal feature sets from spam as well as non-spam data. Then, the classification process
has been performed using Artificial Neural network (ANN) to distinguish the spam and non-spam data. At the end of the process,
performance metrics and comparison will be performed between proposed and existing work to validate the proposed work. The
proposed spam detection system can obtained higher accuracy precision, recall and F-measure compared to the existing
classifiers such as naïve Bayes and Support vector machine (SVM).