Computer technology has advanced to a greater extent which leads to increase in cyber crime committed in recent years. The detection of cyber crime is not an easy task. From the literature, many researchers used various technologies to detect the cyber crime. In this paper, performance evaluation of various techniques are analysed to determine the cyber criminal. Firstly the detection of synthetic identity theft is checked. Secondly, the intrusion detection is checked using the honey pot security mechanism. Thirdly, the detection is further strengthened using the lie detection technique where the false speech of a person is determined. Finally by analysing the user profile, the detection of cyber crime is done using the clustering techniques. Synthetic Identity Theft method performs better than the remaining methods when considered for evaluation. Experimental results show that comparison of the final list of criminal users and the list of criminals determined, the number of genuine users eliminated are 41 out of 100 users, where as the number of genuine users eliminated from other methods are 16, 36 and 38 only. The number of attributes used is only 4, where as the number of attributes used for other methods are 5, 10 and 25. The percentage of performance metrics is also 37.1 and gradient is 31.1 which are better compared to other methods considered for performance analysis.