A novel approach for spam detection based on association rule mining and genetic algorithm. (January 2022)

• Record Type: Journal Article
• Authors: Sokhangoee, Zeynab Fallah; Rezapour, Abdoreza
• Is Part Of: Computers & electrical engineering. Volume 97(2022)
• Publication Date: 2022-01
• Subjects: Spam Detection -- Online Social Network -- Feature Selection -- Genetic Algorithm -- Association Rule Mining; Computer engineering -- Periodicals; Electrical engineering -- Periodicals; Electrical engineering -- Data processing -- Periodicals; Ordinateurs -- Conception et construction -- Périodiques; Électrotechnique -- Périodiques; Électrotechnique -- Informatique -- Périodiques; Computer engineering; Electrical engineering; Electrical engineering -- Data processing; Periodicals; Electronic journals; 621.302854
• DOI: 10.1016/j.compeleceng.2021.107655
• Languages: English
• ISSNs: 0045-7906

Abstract

Highlights: A supervised learning method is proposed for detecting spam in social networks. The proposed method selects the most significant features using Association Rule Mining and Genetic Algorithm. Performance evaluation of proposed method on benchmark datasets gives promising results. Abstract: Spam detection is considered to be one of the most challenging issues in Online Social Networks (OSNs). In this paper, a supervised method is used to detect spam on these platforms. The accuracy of supervised methods depended on two factors: (i) the desired feature selection and (ii) the use of an appropriate classifier. An innovative method is also used for the first factor. This method is a combination of association rule mining and genetic algorithm with the aim of the desired feature selection from a variety of features. On the other hand, the second factor uses a large number of popular classifiers. The proposed method is assessed on three datasets, and the results show the effectiveness of the proposed feature selection method on the accuracy of the classifiers. The average accuracy for both approaches compared to the basic methods is 87.99% and 95.24%, respectively. Graphical abstract: Image, graphical abstract