الدكتور معتز أبو سارةوالأستاذ مراد الزير يشاركان في نشر دراسة بعنوان "تصنيف البريد الإلكتروني من خلال تقنيات تحليل ومعالجة البيانات"

ملخص البحث

This research presents a study of the problem of the difficulty of distinguishing between spam e-mail messages and legitimate e-mail ones. To address this problem, an effective and advanced model for email classification through machine learning algorithms, data analysis, and processing has been proposed and developed. Three powerful algorithms were used: Naïve Bayes, Support Vector Machine (SVM), and Random Forest. While using the first dataset, an accuracy rate of about 97% was observed using the Random Forest algorithm, making it the most accurate algorithm. In comparison, the Naïve Bay algorithm had an accuracy rate of about 95%, while the SVM algorithm had an accuracy rate of 91%. When looking at other metrics, the Random Forest algorithm outperformed the rest. While using the second dataset, an accuracy of about 98% was observed using both the Random Forest and Naïve Bayes algorithms, while the accuracy of the SVM algorithm was about 97%. When looking at other metrics, differences in percentages were observed between the algorithms, as the performance results varied. This confirms the effectiveness of the aforementioned algorithms in email classification, which highlights the importance of using the proposed email classification model, which relies on advanced algorithms and comprehensive data analysis. This model is a powerful solution to the challenges posed by the significant increase in email volume.

كيفية الاستشهاد للبحث:

Zawahra, I., Ashour, W. M., Sara, M. R. A., Zeer, M., & Samhan, A. A. A. (2025). Email Classification Through Data Analysis and Processing Techniques. In From Machine Learning to Artificial Intelligence: The Modern Machine Intelligence Approach for Financial and Economic Inclusion (pp. 641-651). Cham: Springer Nature Switzerland.

رابط البحث

https://link.springer.com/chapter/10.1007/978-3-031-76011-2_44