Article
Telecommunications
P. Rajendran, A. Tamilarasi, R. Mynavathi
Summary: This paper proposes a hybrid approach for collaborative spam detection, which abstracts the entire email layout and extracts layout fingerprints to effectively match and catch the sprouting nature of spam. The system creates a spam database using recommendations from other users, calculates cumulative weights to reduce false positive and false negative ratio, and progressively updates the fingerprints of newly classified spam for up-to-date spam detection. The system is evaluated with the Spam Assassin dataset and shows comparatively better performance.
WIRELESS PERSONAL COMMUNICATIONS
(2022)
Review
Computer Science, Information Systems
L. Earl Gray, Justin M. Conley, Steven J. Bursian, Abu Kamruzzaman, Rameez Asif
Summary: Phishing attacks pose a growing threat to individuals and organizations, and deep learning techniques have the potential to enhance phishing detection. This systematic literature review aims to provide a comprehensive overview of the current state of research on deep learning techniques for phishing detection and identify future research directions.
Article
Engineering, Multidisciplinary
Adel Hamdan Mohammad, Sami Smadi, Tariq Alwada'n
Summary: This research proposes two models for spam detection and feature selection. The first model evaluates the dataset by reducing the number of keywords, and the results are promising. The second model creates features for spam detection and reduces the number of features using three metaheuristic algorithms, with highly significant outcomes.
CMES-COMPUTER MODELING IN ENGINEERING & SCIENCES
(2022)
Article
Computer Science, Theory & Methods
Abdullah Sheneamer
Summary: Email is a cost-effective and efficient method of communication using the internet, with deep learning methods showing promise in effectively filtering out spam emails.
INTERNATIONAL JOURNAL OF ADVANCED COMPUTER SCIENCE AND APPLICATIONS
(2021)
Article
Chemistry, Multidisciplinary
Silvana Gomez-Meire, Cesar Gabriel Marquez, Eliana Patricia Aray-Cappello, Jose R. Mendez
Summary: This study introduces the Live Spam Beater (LiSB) framework for executing email filtering techniques during SMTP transactions, aiming to improve the effectiveness and efficiency of existing proactive filtering mechanisms. By implementing proactive filtering schemes, senders can be notified of spam emails through SMTP response codes during the transaction process. The framework, written in Python, acts as an MTA server and reverse proxy for SMTP, allowing easy incorporation of new proactive filtering techniques through plugins.
APPLIED SCIENCES-BASEL
(2022)
Article
Computer Science, Information Systems
Justinas Rastenis, Simona Ramanauskaite, Ivan Suzdalev, Kornelija Tunaityte, Justinas Janulevicius, Antanas Cenys
Summary: This study presents an automated classification solution based on email message body text, targeting spam and phishing emails. The research explores the limitations of using public datasets to evaluate the necessity of dataset updates for more accurate classification results.
Article
Computer Science, Artificial Intelligence
Wenjuan Li, Lishan Ke, Weizhi Meng, Jinguang Han
Summary: The Internet of Things (IoT) is adopted by many organizations for information collection and sharing. Malicious emails are a security challenge for IoT systems, and email classification using machine learning is a key solution. Empirical research shows that LibSVM and SMO-SVM perform better in email classification.
INTERNATIONAL JOURNAL OF INTELLIGENT SYSTEMS
(2022)
Article
Computer Science, Information Systems
Muhammad Adnan, Muhammad Osama Imam, Muhammad Furqan Javed, Iqbal Murtza
Summary: This study focused on enhancing spam email classification accuracy using stacking ensemble machine learning techniques. The results demonstrated superior performance of the stacking method with the highest accuracy, recall, and F1 score among tested methods. The study presents an innovative combination of classifiers, contributing to the growing body of research on stacking techniques.
INTERNATIONAL JOURNAL OF INFORMATION SECURITY
(2023)
Article
Computer Science, Hardware & Architecture
Federico Concone, Giuseppe Lo Re, Marco Morana, Sajal K. Das
Summary: In recent years, Online Social Networks (OSNs) have revolutionized communication, with platforms like Facebook, Youtube, and Instagram boasting over one billion monthly active users each. Micro-blogging services like Twitter are also popular, with over 120 million users daily sharing global content. Unfortunately, OSNs are plagued by both genuine and malicious users, with the latter spreading unwanted, harmful, and discriminatory content. This article proposes SpADe, a multi-stage spam account detection algorithm that leverages less expensive features initially and extracts complex information only for challenging accounts. Experimental evaluation shows the superiority of this approach over single-stage methods in terms of feature processing and classification time complexity.
IEEE TRANSACTIONS ON DEPENDABLE AND SECURE COMPUTING
(2023)
Article
Computer Science, Information Systems
Maria Novo-Loures, David Ruano-Ordas, Reyes Pavon, Rosalia Laza, Silvana Gomez-Meire, Jose R. Mendez
Summary: This study examines the use of various features to complement synset-based and bag-of-words representations of texts for spam filtering using classical ML approaches. The evaluation of features across different channels and classifiers demonstrates the effectiveness of detecting non-textual entities and using language probability information for classification improvement. Additionally, features influenced by specific behaviors of Internet service users are found to be not useful for spam filtering.
INFORMATION PROCESSING & MANAGEMENT
(2022)
Review
Computer Science, Artificial Intelligence
Francisco Janez-Martino, Rocio Alaiz-Rodriguez, Victor Gonzalez-Castro, Eduardo Fidalgo, Enrique Alegre
Summary: Spam emails are no longer just annoying advertisements, but a growing source of scams and attacks. While machine learning-based spam filters have shown high performance in academic journals, users still face fraudulent and malicious emails. The challenges in this field are the dynamic nature of the environment and the presence of spammers as adversaries.
ARTIFICIAL INTELLIGENCE REVIEW
(2023)
Article
Chemistry, Multidisciplinary
Salman A. Khan, Kashif Iqbal, Nazeeruddin Mohammad, Rehan Akbar, Syed Saad Azhar Ali, Ammar Ahmed Siddiqui
Summary: This paper proposes a new evaluation metric for email spam detection based on fuzzy logic concept, and it confirms the effectiveness through empirical analysis and extrinsic evaluation.
APPLIED SCIENCES-BASEL
(2022)
Article
Computer Science, Artificial Intelligence
Debalina Bera, Obi Ogbanufe, Dan J. Kim
Summary: Despite the presence of anti-phishing filters, social engineering-based cyber-attacks continue to cause significant financial losses, personal identity theft, and loss of sensitive information. This study reviews the literature on psychological attacks in phishing and discusses the need to identify and understand the tactics used by attackers. Using machine learning-based content analysis and topic modeling, the proposed dimensional framework is empirically validated using benchmark datasets of fraudulent emails.
DECISION SUPPORT SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Ghaith Manita, Amit Chhabra, Ouajdi Korbaa
Summary: Given the shortcomings of current methods in spam filtering, we propose an efficient approach called OAOS-LR, which combines an improved AOS algorithm with an LR classification model. By training the LR method with the OAOS approach, the deficiency of low detection rate in the standard LR method is overcome. Experimental results show that OAOS-LR significantly outperforms other methods in spam filtering with high average F1-score success rates on different datasets.
APPLIED SOFT COMPUTING
(2023)
Article
Computer Science, Interdisciplinary Applications
M. Ghiassi, Sean Lee, Swati Ramesh Gaikwad
Summary: This paper introduces a relatively simple and transferrable unsupervised approach to text classification for sentiment analysis and spam filtering. By combining a new clustering algorithm with domain transferrable feature engineering, the integrated solution achieves better accuracy than traditional methods and shows transferability across different datasets.
COMPUTERS & INDUSTRIAL ENGINEERING
(2022)