Deep Self-Supervised Clustering of the Dark Web for Cyber Threat Intelligence
Title | Deep Self-Supervised Clustering of the Dark Web for Cyber Threat Intelligence |
Publication Type | Conference Paper |
Year of Publication | 2020 |
Authors | Kadoguchi, M., Kobayashi, H., Hayashi, S., Otsuka, A., Hashimoto, M. |
Conference Name | 2020 IEEE International Conference on Intelligence and Security Informatics (ISI) |
Date Published | Nov. 2020 |
Publisher | IEEE |
ISBN Number | 978-1-7281-8800-3 |
Keywords | cyber threat intelligence, dark web, deep clustering, Doc2Vec, Human Behavior, human factors, pubcrawl |
Abstract | In recent years, cyberattack techniques have become more and more sophisticated each day. Even if defense measures are taken against cyberattacks, it is difficult to prevent them completely. It can also be said that people can only fight defensively against cyber criminals. To address this situation, it is necessary to predict cyberattacks and take appropriate measures in advance, and the use of intelligence is important to make this possible. In general, many malicious hackers share information and tools that can be used for attacks on the dark web or in the specific communities. Therefore, we assume that a lot of intelligence, including this illegal content exists in cyber space. By using the threat intelligence, detecting attacks in advance and developing active defense is expected these days. However, such intelligence is currently extracted manually. In order to do this more efficiently, we apply machine learning to various forum posts that exist on the dark web, with the aim of extracting forum posts containing threat information. By doing this, we expect that detecting threat information in cyber space in a timely manner will be possible so that the optimal preventive measures will be taken in advance. |
URL | https://ieeexplore.ieee.org/document/9280485 |
DOI | 10.1109/ISI49825.2020.9280485 |
Citation Key | kadoguchi_deep_2020 |