The article “Recognizing phishing websites based on a bayesian combiner” is a research paper that proposes a method for detecting phishing websites using machine learning techniques.
The paper aims to evaluate the performance of different ensemble classifiers, which are models that combine the predictions of multiple base classifiers, such as decision trees, support vector machines, and neural networks.
This paper uses a dataset of 30 features that can distinguish phishing and non-phishing websites, such as the URL length, the presence of HTTPS, and the domain age. The paper employs a stacked generalization strategy, which is a way of combining the outputs of different classifiers using another classifier, called a meta-classifier.
Also, this paper uses three ensemble classifiers as base classifiers: bagging, AdaBoost, and rotation forest. The paper then uses a Bayesian combiner as the meta-classifier, which is a model that calculates the posterior probability of each class given the predictions of the base classifiers. The paper uses 10-fold cross-validation as an evaluation strategy, which is a method of splitting the data into 10 subsets and using each subset as a test set once.
The paper reports the results in terms of accuracy, precision, recall, and F-score, which are metrics that measure the performance of classification models. The paper claims that the proposed method achieves an F-score of 96.3%, which is a good result in detecting phishing websites. The paper concludes that the proposed method can be used as a promising method for identifying phishing websites and suggests some future directions for improving the method. You can read the full article here.
Keywords: visual similarity based approaches, machine learning based solutions, feature extraction techniques, machine learning algorithms, introduction related work