Social network anomaly and attack patterns analysis

The anomaly and attack patterns using data mining have attracted in recent years. The social network anomaly and attack patterns analysis is focused in this research. The data mining system is developed with a cluster algorithm by distance measure and K-Means clustering method to analyze a large amount of IDS log data of social network to discover unknown social network anomaly and attack patterns. The analysis is computed the model and evaluate with test set shows four attackers´ patterns (i) attackers from social network group 2 attempt to execute an arbitrary program on infected systems, (ii) an anomaly pattern came from social network group 1, 5, 6, and 7, (iii) remote attacker from social network group 1 can gain control of vulnerable systems, and (iv) Denial of Service came from social network group 3 and 4. The data mining tools reveal useful insights which are the anomaly and attack patterns. They are applied as a guide to further investigation of social network behaviors.