Title: An efficient intrusion detection system for identification from suspicious URLs using data mining algorithms
Authors: Kotoju Rajitha; Doddapaneni VijayaLakshmi
Addresses: Department of CSE, MGIT, Telangana, India ' Department of CSE, MGIT, Telangana, India
Abstract: The main objective of this paper is to design intrusion detection from suspicious URLs using optimal fuzzy logic system. Basically, the system consists of three modules such as: 1) feature extraction; 2) feature selection; 3) classification. At first, we extract the four kinds of feature from the dataset which have a total of 30 features. Among that, we select the important features using hybridisation of firefly and cuckoo search algorithm (HFFCS). Then, we train the selected features using fuzzy logic classifier and then we calculate the fuzzy logic score. Finally in testing, the fuzzy logic classifier detected the malicious URL based on the fuzzy score. In this work, we use two types of database such as URL reputation dataset and phishing websites dataset. The experimental results demonstrate that the proposed malicious URL detection method outperforms other existing methods.
Keywords: malicious; uniform resource locator; URL; detection; firefly; cuckoo search; fuzzy logic classifier; FLC; websites detection.
DOI: 10.1504/IJBIDM.2017.084284
International Journal of Business Intelligence and Data Mining, 2017 Vol.12 No.2, pp.133 - 158
Received: 30 Jul 2016
Accepted: 01 Dec 2016
Published online: 23 May 2017 *