基于深度学习的实体关系抽取方法研究

An entity relation extraction method based on deep learning

常用的神经网络,如卷积神经网络(CNN)和循环神经网络(RNN),在关系抽取任务中都表现出了很不错的效果。然而,卷积神经网络擅长捕获局部特征,但不太适合处理序列特征;传统的循环神经网络虽然可以有效提取长距离词之间的特征,但容易出现梯度消失或梯度爆炸问题。针对这些问题,提出了一种结合BiLSTM-CNN-Attention的混合神经网络模型。BiLSTM和CNN的结合使它们优劣互补,而Attention的引入能够突出实体间关系词在整个句子中的重要性。并且,在词嵌入层使用拼接词向量,克服了词向量单一表示的问题。实验结果表明,相比word2vec词向量,拼接词向量能够获取语义更丰富的词向量,使词向量的健壮性更强。与BiLSTM-CNN、CNN-Attention和BiLSTM-Attention模型相比,BiLSTM-CNN-Attention混合模型的准确率和F1值都有所提升。

Commonly used neural networks such as Convolutional Neural Network(CNN)and Recurrent Neural Network(RNN)have shown very good results in relation extraction tasks.However,CNN is good at capturing local features,but it is not suitable for processing sequence features.Traditional RNN can effectively extract features between long-distance words,but it is easy to cause gradient disappearance or gradient explosion.To solve these problems,a hybrid neural network model,called BiLSTM-CNN-Attention,is proposed.The combination of BiLSTM and CNN makes them complement each other,and the introduction of Attention can highlight the importance of inter entity relation words in the whole sentence.In addition,the mosaic word vector is used in the word embedding layer to overcome the problem of single word vector representation.The experimental results show that,compared with word2vec word vector,mosaic word vector can obtain more semantic word vector and enhance the robustness of word vector.Compared with BiLSTM-CNN,CNN-Attention and BiLSTM-Attention models,BiLSTM-CNN-Attention improves the accuracy and F1 value.