基于LSTM神经网络的SQL注入攻击检测研究  

Research on SQL injection attack detection based on LSTM neural network

作　　者：楚翔皓 刘震 CHU Xiang-hao;LIU Zhen(School of Computer Science and Engineering,Tianjin University of Technology,Tianjin 300384,China;School of General Information Science,Nagasaki Institute of Applied Science,Nagasaki 851-0193,Japan)

机构地区：[1]天津理工大学计算机科学与工程学院,天津300384 [2]长崎综合科学大学综合信息学部,长崎日本851-0193

出　　处：《天津理工大学学报》2019年第6期41-46,共6页Journal of Tianjin University of Technology

摘　　要：由于Web技术的不断发展,尤其是5G技术的逐渐普及,万物联网的时代已经到来.网络的便利让数据的流通更加的及时和有价值,各种Web应用不仅方便了人们的日常生活,推进了社会的进步,更是带来了巨大的经济效益.因此,许多不法分子以攻击他人Web应用的方式来获取盈利,而作为OWASP(Open WebApplication Security Project)公布的报告中排名第一的注入类漏洞,是不法分子最常攻击的方向,带来了巨大的安全威胁.近年来,众多安全研究者都着力于找寻出更高效,准确度更高的方式来防御SQL注入攻击,本文将会对当前研究状况进行总结分析,并提出一种新基于LSTM神经网络的SQL注入语句分类识别模型.Due to the continuous development of Web technologies,especially the gradual popularization of 5G technology,the era of the Internet of Things has arrived.The convenience of the network makes the circulation of data more timely and valuable.Various network applications not only facilitate people's daily life,but also promote the progress of society and bring huge economic benefits.As the top one injection class vulnerability in the report published by OWASP(Open Web Application Security Project),is the most common direction of attack by criminals,and has brought a huge security threat.In recent years,many security researchers have focused on finding more efficient and accurate ways to defend against SQL injection attacks.This paper will summarize the current research situation and propose a new SQL injection statement classification recognition model based on LSTM neural network.

关 键 词：SQL注入 Long Short-Term Memory(LSTM)神经网络 机器学习 特征提取 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]