Knowledge Graph creation on Windows malwares and completion using knowledge graph embedding
Omkar Shinde, Anish Khobragade · 2022
In current internet era, it is convenient to access information within few clicks. This ease of access to information is not only attracting more consumer markets but also attracts various computer threats. The Microsoft windows is the most popular desktop operating system used worldwide due to its simplicity, various hardware vendor’s support. Such massive user base attracts various hacker’s attention to exploit vulnerabilities present in windows. Windows Portable executable is file format used by windows to store executable files. Portable Executable file contain various sections, among those import section contain OS functionality required by executable file for it’s working. This work proposes Knowledge graph from real windows malware executables and tries to complete it using knowledge graph embedding models. Live Malware executables are gathered from Dike dataset and entity, relations are extracted accordingly. Later Knowledge graph embedding models are used for completion and various models are evaluated based on metrices like hit@k, MR, MRR.