结合注意力机制与双向LSTM的中文事件检测方法
首发时间:2019-03-18
摘要:事件检测是信息抽取领域的重要任务之一。已有的方法大多高度依赖于复杂的语言特征工程和自然语言处理工具,中文事件检测还存在着由分词带来的触发词分割问题。本文将中文事件检测视为一个序列标注而非分类问题,提出了一种结合注意力机制与长短期记忆神经网络的中文事件检测模型ATT-BiLSTM,利用注意力机制来更好地捕获全局特征,并通过两个双向LSTM层来更有效地捕获句子序列特征,从而提高中文事件检测的效果。在ACE 2005中文数据集上的实验表明,本文提出的方法与其他现有的中文事件检测方法相比性能得到了明显的提升。
For information in English, please click here
Chinese Event Detection Method Combining Attention Mechanism and BiLSTM
Abstract:Event detection is one of the important tasks in the field of information extraction. Most of the existing methods are highly dependent on complicated natural language feature engineering and processing tools. Chinese event detection even suffers from the problem that some trigger words are segmented by word segmentation. In this paper, Chinese event detection is regarded as a sequence labeling rather than a classification problem. A Chinese event detection model ATT-BiLSTM is proposed, which integrates attention mechanism and long short-term memory neural network. The attention mechanism is used to better capture global features and BiLSTM layers are employed to capture sequence features more effectively, which can improve the performance of Chinese event detection. Experiments on the ACE 2005 Chinese dataset show that the performance of the proposed method significantly outperforms other existing Chinese event detection methods.
Keywords: Chinese event detection attention mechanism long short-term memory
引用
No.****
动态公开评议
共计0人参与
勘误表
结合注意力机制与双向LSTM的中文事件检测方法
评论
全部评论0/1000