USING METADATA IN FILTERED LOGS FOR PREVENTION OF DATABASE INTRUSION THROUGH OWNER LEVEL INTERACTION

As the online activities increases, the chance for attacks also increases in a rapid manner. Among the different categories of attacks, database attacks are most popular, since the data residing on it are highly confidential. One of the main obstacles that we face while conducting forensics analysis is the huge amount of data that we need to analyze in which the required amount of information will be very less and the lack of evidence regarding the intrusion. So in this paper, we focus on a method to filter the bulk amount of log data using a reduction algorithm based on frequent attribute values. Then using metadata, evidence regarding the intrusion is retrieved and notifies the database user about the attempt to mitigate its effect by reversing the act ion performed by the intruder.