With every passing second social network community is growing rapidly, because of that, attackers have shown keen interest in these kinds of platforms and want to distribute mischievous contents on these platforms. With the focus on introducing new set of characteristics and features for counteractive measures, a great deal of studies has researched the possibility of lessening the malicious activities on social media networks. This research was to highlight features for identifying spammers on Instagram and additional features were presented to improve the performance of different machine learning algorithms. Performance of different machine learning algorithms namely, Multilayer Perceptron, Random Forest, K-Nearest Neighbor and Support Vector Machine were evaluated on machine learning tools named, RapidMiner and WEKA. The result from this research tells us that Random Forest outperformed all other selected machine learning algorithms on both selected machine learning tools. Overall, Random Forest provided best results on RapidMiner. These results are useful for the researchers who are keen to build machine learning models to find out the spamming activities on social network communities.

Keywords : Malicious Activities, Spammers, Machine Learning Algorithms, Multilayer Perceptron, Random Forest, K-Nearest Neighbor, Support Vector Machine, Rapidminer, WEKA, Social Network Communities.