Scholarlink

With the development of various technologies, the network security environment is facing more and more threats, and these new security threats tend to be more insidious, which makes it difficult for traditional analysis methods to respond effectively. Currently, people are facing data leakage, ransomware attacks, advanced persistent attacks, and other network security problems, and they need effective detection means to accurately detect the behavior of malicious code. Static analysis of the malicious code from the assembly code API call instructions according to the regularization of the expression to extract API call sequences. The extracted API sequences will be embedded into the classification model GRU with the Doc2vec model. GRU model is built for API call classification. The channel attention mechanism is introduced to help the model to capture the key features of malicious code. Capture the critical features of malicious code. Finally, the multi-classification results are evaluated using loss function and accuracy metrics.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

An Approach for Malware Classification via API Call Extraction from Disassembly Code

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Jian Chen,
  • Zhibin Yang,
  • Guangao Li,
  • Jun Wang,
  • Siluo Sun,
  • Shoudao Sun,
  • Yuntao Zhao,
  • Yuerong Li

摘要

With the development of various technologies, the network security environment is facing more and more threats, and these new security threats tend to be more insidious, which makes it difficult for traditional analysis methods to respond effectively. Currently, people are facing data leakage, ransomware attacks, advanced persistent attacks, and other network security problems, and they need effective detection means to accurately detect the behavior of malicious code. Static analysis of the malicious code from the assembly code API call instructions according to the regularization of the expression to extract API call sequences. The extracted API sequences will be embedded into the classification model GRU with the Doc2vec model. GRU model is built for API call classification. The channel attention mechanism is introduced to help the model to capture the key features of malicious code. Capture the critical features of malicious code. Finally, the multi-classification results are evaluated using loss function and accuracy metrics.