An Efficient Partition Technique to reduce the Attack Detection Time with Web based Text and PDF files

In this paper we propose an efficient partition technique for web based files (jsp, html, php), text (word, text files) and PDF files. We are working in the direction of attack time detection. For this motivation we are considering mainly two factors first in the direction of minimizing the time, second in the direction of file support. For minimizing the time we use partitioning method. We also apply partitioning method on PDF files. In this paper we proposed an efficient hybrid approach which is the combination of Key generation, partitioning and encry ption. In this approach admin register the client after the connection request from the client. Then client can approach for the data. Admin now call the data preparation strategy for data preparation for sending. In the preparation s tage admin first gener ate the key which is common for server and client, then partition the data, so that the file overhead is reduced and encrypt the file by java default encryption method. After successfully preparation, the data log will be maintained and the data will be se nding to the client by adding a hidden bit. When any attacker updates the data the hidden bit is automatically changes and an alert will be generated to the server. Server record the time of alert and inform to the client by changing the hidden bit, client understands the attack by checking their log file. If the hidden bit is changed in the client side then client understand the attack and send the request again for the same file. In the result section we provide the comparison with the traditional techniq ue which shows the effectiveness of our approach.