BERT-GraphSAGE: hybrid approach to spam detection
Abstract Even after the advent of various communication networks, emails have retained their importance and serious, professional character. Moreover, as the number of Internet users increases, so does the number of spam emails. Spam refers to any unsolicited and unwanted communication, which leads...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
SpringerOpen
2025-05-01
|
| Series: | Journal of Big Data |
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s40537-025-01176-9 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850268947315687424 |
|---|---|
| author | F. Zouak O. El Beqqali J. Riffi |
| author_facet | F. Zouak O. El Beqqali J. Riffi |
| author_sort | F. Zouak |
| collection | DOAJ |
| description | Abstract Even after the advent of various communication networks, emails have retained their importance and serious, professional character. Moreover, as the number of Internet users increases, so does the number of spam emails. Spam refers to any unsolicited and unwanted communication, which leads to a significant waste of resources and overloads networks. The majority of these spam emails come from advertisers wishing to promote their products, while others have more malicious intentions, such as phishing emails aimed at tricking recipients into providing confidential information like website credentials or credit card details. In our research, we aim to improve spam detection by combining BERT (Bidirectional Encoder Representations from Transformers) and GraphSAGE. The embedding vectors generated by BERT are used to represent the nodes of a graph, which are then linked based on the calculation of cosine similarity. This graph structure is subsequently exploited by GraphSAGE, which doesn’t merely record the results of embedding mapping; it learns an inductive method of embedding generation. This enables GraphSAGE to generalize to unseen emails by sampling and aggregating the characteristics of neighboring emails to produce robust node representations. Our model was evaluated on three benchmark datasets: ENRON, SpamAssassin, and LingSpam. It achieved 98.87% accuracy, 99.81% precision, and 99.98% AUC on ENRON, 96.44% accuracy, 94.43% precision, and 98.86% AUC on SpamAssassin, and 99.20% accuracy, 96.98% precision, and 99.55% AUC on LingSpam, outperforming several state-of-the-art baselines. These results confirm the robustness of our approach in accurately distinguishing between spam and legitimate emails. |
| format | Article |
| id | doaj-art-5751dc205f4c460faf2d01c67ac7b418 |
| institution | OA Journals |
| issn | 2196-1115 |
| language | English |
| publishDate | 2025-05-01 |
| publisher | SpringerOpen |
| record_format | Article |
| series | Journal of Big Data |
| spelling | doaj-art-5751dc205f4c460faf2d01c67ac7b4182025-08-20T01:53:19ZengSpringerOpenJournal of Big Data2196-11152025-05-0112111410.1186/s40537-025-01176-9BERT-GraphSAGE: hybrid approach to spam detectionF. Zouak0O. El Beqqali1J. Riffi2LISAC Laboratory, Faculty of Sciences Dhar El Mahraz, University Sidi Mohamed Ben AbdellahLISAC Laboratory, Faculty of Sciences Dhar El Mahraz, University Sidi Mohamed Ben AbdellahLISAC Laboratory, Faculty of Sciences Dhar El Mahraz, University Sidi Mohamed Ben AbdellahAbstract Even after the advent of various communication networks, emails have retained their importance and serious, professional character. Moreover, as the number of Internet users increases, so does the number of spam emails. Spam refers to any unsolicited and unwanted communication, which leads to a significant waste of resources and overloads networks. The majority of these spam emails come from advertisers wishing to promote their products, while others have more malicious intentions, such as phishing emails aimed at tricking recipients into providing confidential information like website credentials or credit card details. In our research, we aim to improve spam detection by combining BERT (Bidirectional Encoder Representations from Transformers) and GraphSAGE. The embedding vectors generated by BERT are used to represent the nodes of a graph, which are then linked based on the calculation of cosine similarity. This graph structure is subsequently exploited by GraphSAGE, which doesn’t merely record the results of embedding mapping; it learns an inductive method of embedding generation. This enables GraphSAGE to generalize to unseen emails by sampling and aggregating the characteristics of neighboring emails to produce robust node representations. Our model was evaluated on three benchmark datasets: ENRON, SpamAssassin, and LingSpam. It achieved 98.87% accuracy, 99.81% precision, and 99.98% AUC on ENRON, 96.44% accuracy, 94.43% precision, and 98.86% AUC on SpamAssassin, and 99.20% accuracy, 96.98% precision, and 99.55% AUC on LingSpam, outperforming several state-of-the-art baselines. These results confirm the robustness of our approach in accurately distinguishing between spam and legitimate emails.https://doi.org/10.1186/s40537-025-01176-9BERTCosine similarityGraphSAGENode classificationSpam detection |
| spellingShingle | F. Zouak O. El Beqqali J. Riffi BERT-GraphSAGE: hybrid approach to spam detection Journal of Big Data BERT Cosine similarity GraphSAGE Node classification Spam detection |
| title | BERT-GraphSAGE: hybrid approach to spam detection |
| title_full | BERT-GraphSAGE: hybrid approach to spam detection |
| title_fullStr | BERT-GraphSAGE: hybrid approach to spam detection |
| title_full_unstemmed | BERT-GraphSAGE: hybrid approach to spam detection |
| title_short | BERT-GraphSAGE: hybrid approach to spam detection |
| title_sort | bert graphsage hybrid approach to spam detection |
| topic | BERT Cosine similarity GraphSAGE Node classification Spam detection |
| url | https://doi.org/10.1186/s40537-025-01176-9 |
| work_keys_str_mv | AT fzouak bertgraphsagehybridapproachtospamdetection AT oelbeqqali bertgraphsagehybridapproachtospamdetection AT jriffi bertgraphsagehybridapproachtospamdetection |