Accession Number : AD1025911


Title :   Intrusion Detection Systems with Live Knowledge System


Descriptive Note : Technical Report,20 May 2015,19 May 2016


Corporate Author : UNIVERSITY OF TASMANIA SANDY BAY Australia


Personal Author(s) : Kang,Byeong H


Full Text : https://apps.dtic.mil/dtic/tr/fulltext/u2/1025911.pdf


Report Date : 31 May 2016


Pagination or Media Count : 10


Abstract : Detecting phishing websites has been noted as a complex and dynamic problem area because of the subjective considerations and ambiguities of detection mechanism. Either machine learning technique or human expert system has been applied to acquire and maintain the knowledge for phishing website detection and prediction but neither did work successfully. In this project, we propose novel approach that uses Ripple-down Rule (RDR) to maintain the knowledge from human experts with knowledge base generated by the Induct RDR, which is a machine-learning based RDR algorithm. The performance of proposed model was compared with that of 6 different machine-learning techniques. Our experimental results showed the proposing approach can help to deduct the cost of solving over-generalization and overfitting problems of machine learning approach.


Descriptors :   machine learning , artificial intelligence software , knowledge based systems , artificial intelligence computing , artificial neural networks , expert systems , electronic mail , algorithms , intrusion detection systems


Subject Categories : Computer Systems Management and Standards
      Computer Programming and Software


Distribution Statement : APPROVED FOR PUBLIC RELEASE