A Survey on Log Mining: A Data Mining Approach for Intrusion Detection

There are many approaches present in today’s world to protect data as well as network. One of the way is an Intrusion Detection System (IDS) to make data more secure. Many researches are done in the field of intrusion detection, but the main concentration of these researches is on the networks and operating system. The unauthorized access may lead to break the integrity of the system as it may be in the form of execution of malicious transaction. E-commerce is one of the sectors suffers from million dollar losses only because of these unauthorized activities and malicious transactions. So, it is today’s demand to detect malicious transactions and also to provide some protection. In this paper, we provided the detection system for intrusion detection in the e-commerce system and we are also trying to avoid different types of attack by applying different preventive measures. For detecting malicious transactions,we are going to use one of the data mining algorithm weighted data dependency rule minerfor our eCommerce database IDS. Which, extracts the read-write dependency rules to check whether the transactions are malicious or not. This system finds the malicious transactions as well as identify the transactions that performs read write operations without permission.