基于深度学习理论的电子商务商品实体智能识别

E-commerce Commodity Entity Intelligent Identification Based on Deep Learning Theory

为了提升电子商务商品实体智能识别的准确性,提出基于深度学习理论的电子商务商品实体智能识别方法。首先将电子商务商品文本序列的各个字词转换为字词向量,并作为深度学习网络输入,然后深度学习网络根据上下文信息获取特征向量隐藏序列,产生字词向量矩阵,最后采用Self-Attention映射字词向量矩阵,训练获取权重矩阵,得到分数矩阵,并通过条件随机设置标注限制,输出电子商务商品命名实体识别结果。实验结果表明,深度学习理论可以提高电子商务商品实体智能识别效果,具有较高的实际应用价值。

In order to improve the accuracy of e-commerce commodity entity intelligent recognition,this paper proposes an e-commerce commodity entity intelligent recognition method based on deep learning theory.Firstly,each word of e-commerce commodity text sequence is transformed into word vector,which is used as the input of deep learning network.Then,the hidden sequence of feature vector is obtained by deep learning network according to the context information,and the word vector matrix is generated.Finally,the weight matrix is obtained by training and the score matrix is obtained by training using Self-Attention mapping word vector matrix set label restriction.It outputs e-commerce commodity named entity recognition results.The experimental results show that the deep learning theory can improve the effect of e-commerce commodity entity intelligent recognition,and has high practical application value.