信息网络安全 ›› 2017, Vol. 17 ›› Issue (9): 153-156.doi: 10.3969/j.issn.1671-1122.2017.09.035

• • 上一篇    下一篇

基于大数据样本的软件行为安全分析

郭敏(), 曾颖明, 姚金利, 达小文   

  1. 北京计算机技术及应用研究所, 北京 100854
  • 收稿日期:2017-08-01 出版日期:2017-09-20 发布日期:2020-05-12
  • 作者简介:

    作者简介: 郭敏(1991—),女,山东,助理工程师,硕士,主要研究方向为信息安全;曾颖明(1985—),男,江西,高级工程师,硕士,主要研究方向为信息安全;姚金利(1984—),男,山西,工程师,硕士,主要研究方向为信息安全;达小文(1989—),男,湖南,助理工程师,硕士,主要研究方向为信息安全。

The Analysis of Software Behavior Security Based on Big Data Samples

Min GUO(), Yingming ZENG, Jinli YAO, Xiaowen DA   

  1. Beijing Institute of Computer Technology and Applications, Beijing 100854, China
  • Received:2017-08-01 Online:2017-09-20 Published:2020-05-12

摘要:

因软件的不当行为或恶意破坏导致的信息系统被攻击事件频发,信息系统对软件的安全性要求越来越高,如何有效实现对软件行为的安全性分析成为业界的一个研究热点。文章重点研究了基于大数据样本的软件行为分析技术,针对大数据样本海量、多维度、高速多变、内部关联关系复杂等特点,采用基于层次聚类算法的静态分析和基于SVM算法的动态行为分析相结合的方法,构建基于机器学习算法的软件行为安全分析模型。该模型采用云端集中处理的方法,可有效节省终端的资源消耗,实现对恶意软件的高效、快速检测。

关键词: 大数据样本, 软件行为, 机器学习算法

Abstract:

Because information system attack events caused by software misconducts or malicious damages occur frequently-software security requirements of information system are higher and higher. How to achieve the security analysis of software behaviors effectively has become a popular topic. This paper focuses on the software behaviors analysis technology based on big data samples. Considering the characteristics of massive, multi-dimensional, high-speed change and complex internal relations of big data samples, combining the static analysis based on hierarchical clustering algorithm with dynamic behaviors analysis based on SVM algorithm, this paper constructs a software behaviors analysis model based on machine learning algorithm. The model uses cloud centralized processing method, which can effectively save the resource consumption of the terminal, and realize the efficient and rapid detection of malicious software.

Key words: big data samples, software behavior, machine learning algorithm

中图分类号: