Fortification of automated network systems from intrusive attacks is one of the major issues in the field of network security. The proposed work dealt with the process of identifying false positive attacks from various sources of network. In this anticipated research work we have predicted the decisive class label of a network packet for either accept or reject the request and this can be achieved by applying various data mining approaches with a novel steganographic approach, which helps to attain the prediction process. The simulation tests are conducted using KDD Cup 99 dataset and Arnasa Web Third International Knowledge Discovery and Data Mining Tools Competition and Medical consultation sity of the Basque Country. In network system environment the packet pattern is evaluated using signature and the packet is either accepted or rejected. The proposed methodology works on the Intrusion Detection System application and predicts the new requests on the basis of the existing knowledge and the identified knowledge is regularly updated in the Knowledge Base. The results show that the overall prediction accuracy for the occurrence and nonoccurrence of the false positive attacks on the network environment using the mining procedure is acceptable.