| dc.contributor | 資管系 | |
| dc.creator (作者) | 蕭舜文 | |
| dc.creator (作者) | Hsiao, Shun-Wen;Kuok, Kelvin Io Wai | |
| dc.date (日期) | 2025-10 | |
| dc.date.accessioned | 9-Dec-2025 10:42:59 (UTC+8) | - |
| dc.date.available | 9-Dec-2025 10:42:59 (UTC+8) | - |
| dc.date.issued (上傳時間) | 9-Dec-2025 10:42:59 (UTC+8) | - |
| dc.identifier.uri (URI) | https://ah.lib.nccu.edu.tw/item?item_id=180144 | - |
| dc.description.abstract (摘要) | Cybercrime costs are increasing, generating large amounts of adversarial data captured by honeypots. Manual analysis is impractical, and non-content-based machine learning methods are unsuitable for understanding attack intention. This research presents a content-based intelligent system designed to analyze network packets collected via honeypots. The system utilizes a BERT language model pre-trained on network packets to transform HTTP packet content into 768-dimensional vector, capturing semantic and syntactic information. These vectors enable downstream applications such as unsupervised clustering to group similar attack patterns. This approach allows for efficient processing of large malicious packets and provides a deep understanding of malicious activity characteristics and intentions. | |
| dc.format.extent | 110 bytes | - |
| dc.format.mimetype | text/html | - |
| dc.relation (關聯) | IEEE Conference on Dependable and Secure Computing (IEEE DSC) 2025, IEEE | |
| dc.subject (關鍵詞) | honeypot; language model; packet; HTTP | |
| dc.title (題名) | HTTP Adversarial Activity in Honeypots | |
| dc.type (資料類型) | conference | |
| dc.identifier.doi (DOI) | 10.1109/DSC65356.2025.11260888 | |
| dc.doi.uri (DOI) | https://doi.org/10.1109/DSC65356.2025.11260888 | |
