Visible to the public Research on Filtering Feature Selection Methods for E-Mail Spam Detection by Applying K-NN Classifier

TitleResearch on Filtering Feature Selection Methods for E-Mail Spam Detection by Applying K-NN Classifier
Publication TypeConference Paper
Year of Publication2022
AuthorsGeorgieva-Trifonova, Tsvetanka
Conference Name2022 International Congress on Human-Computer Interaction, Optimization and Robotic Applications (HORA)
Date Publishedjun
KeywordsEuclidean distance, feature extraction, feature selection, Filtering, Human Behavior, human computer interaction, K-NN classifier, Metrics, pubcrawl, robots, Scalability, spam detection, text categorization, unsolicited e-mail
AbstractIn the present paper, the application of filtering methods to select features when detecting email spam using the K-NN classifier is examined. The experiments include computation of the accuracy and F-measure of the e-mail texts classification with different methods for feature selection, different number of selected features and two ways to find the distance between dataset examples when executing K-NN classifier - Euclidean distance and cosine similarity. The obtained results are summarized and analyzed.
DOI10.1109/HORA55278.2022.9799999
Citation Keygeorgieva-trifonova_research_2022