基于大數(shù)據(jù)分析的網(wǎng)絡(luò)安全威脅檢測(cè)系統(tǒng)研究

打開(kāi)文本圖片集

doi：10.3969/J.ISSN.1672-7274.2025.06.017

中圖分類(lèi)號(hào)：TP3 文獻(xiàn)標(biāo)志碼：B 文章編碼：1672-7274（2025）06-0050-04

Research on Network Security Threat Detection System Based on Big Data Analysis

ZHANGLianlian

（Chifeng Vocational College of Applied Technology，Chifeng O24oo5，China）

Abstract： With the increasing complexity of network security threats， traditional detection methods are facing challenges.This article proposes a network security threat detection system that combines big data analysis and deep learning.The system uses Apache Spark to process network traffcand log data，extract features，and improve the accuracyand real-time performance of network security threat detection through a combined model of LSTM and XGBoost.The experimental results show that compared to traditional SVM and decision tres， the proposed method has improved accuracy by 4.4% and recall rate by 10.7% （compared to SVM）. Although there is a certain detection delay，it stillmeets the real-time detectionrequirements.Research has shown thatcombining big data and dep learning models can efectively improve network security threat detection capabilities and provide new ideas for network security protection.

Keywords：security threat detection;deep learning;bigdataanalysis

1 研究背景

隨著網(wǎng)絡(luò)攻擊技術(shù)不斷升級(jí)，傳統(tǒng)安全措施難以應(yīng)對(duì)復(fù)雜的網(wǎng)絡(luò)威脅。（剩余4487字）