摘要
随着互联网的普及,钓鱼网站是人们接触到最常见的网络犯罪类型之一,造成了巨大的经济损失、信息泄漏等危害。鉴于此,提出了一种基于双向长短记忆神经网络(Bi-LSTM)结合注意力机制模型来识别钓鱼网站。预处理之后的数据,采用Word2Vec模型构建词向量,通过Bi-LSTM进行特征提取,并使用注意力机制计算注意力权重,生成最终的特征向量,通过Sigmoid函数对网站类别分类。实验结果表明,所提出的模型在对钓鱼网站检测效果达到98.3%,能够有效应对此类攻击威胁,有助于维护网络空间安全体系。
With the popularization of the Internet,phishing websites that people contact are one of the most common types of cybercrime,causing huge economic loss and information leakage.Based on this,a model based on Bi-LSTM(bidirectional long short memory neural network)combined with an attention mechanism is proposed to recognize phishing websites.Data were prepro-cessed,word vectors were constructed using the Word2Vec model,features is extracted using Bi-LSTM,attention weights were cal-culated using the attention mechanism to generate final feature vectors,and the classification of phishing websites was obtained by using the sigmoid function.Experimental results show that the model is 98.3%effective in detecting phishing sites,and can effec-tively deal with the threat of phishing attacks and maintain the security of phishing sites.
作者
尚培文
李东帅
Shang Peiwen;Li Dongshuai(School of Electronics&Information Engineering,Liaoning University of Technology,Jinzhou 121001,China)
出处
《现代计算机》
2024年第6期50-55,共6页
Modern Computer
