Prediksi Website Pemancing Informasi Penting Phising Menggunakan Support Vector Machine (SVM)
Abstract
Abstrak: Perkembangan teknologi informasi dan komunikasi khususnya internet berdampak pada semua sektor kehidupan manusia tidak terkecuali dengan sektor perbankan dan keuangan. Selain memberikan dampak positif dengan dipermudahnya pelanggan dalam proses transaksi yang dapat dilakukan kapanpun dan di manapun tanpa dibatasi oleh ruang dan waktu menggunakan media internet, juga membawa potensi besar terhadap pihak-pihak yang tak bertanggungjawab untuk melakukan pencurian data dan informasi penting, salah satunya dengan teknik phishing, sehingga metode untuk mendeteksi serangan situs phishing memerlukan perhatian serius. Dalam penelitian ini penulis telah melakukan memberikan gambaran metode yang paling akurat untuk mendeteksi website phishing dengan membandingkan tiga metode antara lain Support Vector Machine, Naïve Bayes, dan Decision Tree menggunakan dataset publik dari UCI Machine Learning Repository (www.uci.edu) yang dioptimasi dengan feature selection dan diolah menggunakan program RapidMiner. Hasil penelitian menunjukan bahwa metode Decision Tree mempunyai tingkat akurasi sebesar 91,84%, metode Naïve Bayes sebesar 74,07% dan Support Vector Machine sebesar 92,34%. Hal ini menunjukan bahwa metode Support Vector Machine mempunyai tingkat akurasi yang paling tinggi..Kata Kunci: Decision Tree, Naïve Bayes, Phishing, Support Vector Machine
Abstract: The development of information and communication technologies, especially the Internet, have an impact in all sectors of human life with exception in the banking and financial sectors in addition to a positive impact to make essier customer in the transaction process that can do anytime and anywhere without being limited by space and time using the internet, it also brings great potential against parties not responsible for the theft of critical data and information, one of them with phishing techniques, so the method for detecting a phishing site requires serious attention. In this study the authors try to give an overview of the most accurate methods to detect phishing websites to compare three methods such as Support Vector Machine, Naïve Bayes, and Decision Tree using public datasets from the UCI Machine Learning Repository (www.uci.edu) optimized with feature selection and processed using RapidMiner program that showed Decision Tree has a accuracy rate of 91.84%, Naïve Bayes method amounted to 74.07% and Support Vector Machine by 92.34%. Hereby declare that the method of Support Vector Machine has the highest degree of accuracy.
Keyword: Decision Tree, Naïve Bayes, Phishing, Support Vector Machine
References
Bhanji A, Jadhav P, Bhujbal S, Mulak P, Phishing K-, Introduction I. 2013. ER ER. 2: 2340–2347.
Chunjiang H, Cuilian Z, Yan Z. 2009. A New SVM Merged into Data Information. 2009 Asia-Pacific Conf. Inf. Process. I: 14–17.
Han J, Rodriguze JC, Beheshti M. 2008. Diabetes Data Analysis and Prediction Model Discovery Using RapidMiner.
James L. 2005. Phising Exposed. Stewart J, editor. United States. 1-382 p.
Lin S, Shiue Y, Chen S, Cheng H. 2009. Expert Systems with Applications Applying enhanced data mining approaches in predicting bank performance : A case of Taiwanese commercial banks. 36: 11543–11551.
Liu Y. 2011. An adaptive fuzzy ant colony optimization for feature selection An Adaptive Fuzzy Ant Colony Optimization for Feature Selection. 1–8.
Long J. 2008. No Tech Hacking: A Guide to Social Engineering, Dumpster Diving, and Shoulder Surfing. Pinzon Scott, editor. United States: Andrew Williams. 1-285 p.
Maimon O, Rokach L. 2010. Data Mining and Knowledge Discovery Handbook, Second. Rokach L, editor. 21-36 p.
Martino AS, Perramon X. 2010. Phishing Secrets : History , Effects , and Countermeasures. 11: 163–171.
Vapnik VN. 1999. An Overview of Statistical Learning Theory. 10: 988–999.
Vercellis C. 2009. Business Intelligence: Data Mining and Optimization for Decision Making. Italy. 1-417 p.
Weiss S. 2010. Text Mining : Predictive Methods for Analysis and Prediction Model Discovery Using RapidMiner. Indurkhya, editor. New Jersey: Springer Science & Business Media. 1-237 p.
Zhao M, Fu C, Ji L, Tang K, Zhou M. 2011. Expert Systems with Applications Feature selection and parameter optimization for support vector machines : A new approach based on genetic algorithm with feature chromosomes. 38: 5197–5204.
