基于神经网络的连动句识别  

Recognition of Serial-verb Sentences Based on Neural Network

作　　者：孙超[1,4] 曲维光 魏庭新[2,3] 顾彦慧 李斌[2] 周俊生[1] SUN Chao;QU Weiguang;WEI Tingxin;GU Yanhui;LI Bin;ZHOU Junsheng(School of Computer and Electronic Information,Nanjing Normal University,Nanjing,Jiangsu 210023,China;School of Chinese Language and Literature,Nanjing Normal University,Nanjing,Jiangsu 210097,China;International College for Chinese Studies,Nanjing Normal University,Nanjing,Jiangsu 210097,China;Kunshan Zhenchuan Senior High School,Suzhou,Jiangsu 215300,China)

机构地区：[1]南京师范大学计算机与电子信息学院,江苏南京210023 [2]南京师范大学文学院,江苏南京210097 [3]南京师范大学国际文化教育学院,江苏南京210097 [4]昆山震川高级中学,江苏苏州215300

出　　处：《中文信息学报》2022年第2期12-21,共10页Journal of Chinese Information Processing

基　　金：国家自然科学基金(61772278);国家社会科学基金(21&ZD288,18BYY127);教育部人文社会科学研究青年基金(17YJC740084)。

摘　　要：连动句是具备连动结构的句子,是汉语中一种特殊的句法结构,在现代汉语中十分常见且使用频繁。连动句语法结构和语义关系都很复杂,在识别中存在许多问题,对此该文针对连动句的识别问题进行了研究,提出了一种基于神经网络的连动句识别方法。该方法分两步:第一步,运用简单的规则对语料进行预处理;第二步,利用文本分类的思想,使用BERT编码,利用多层CNN与BiLSTM模型联合提取特征进行分类,进而完成连动句识别任务。在人工标注的语料上进行实验,实验结果达到92.71%的准确率,F_(1)值为87.41%。Serial-verb sentence is a sentence with several coordinated verbs.The grammatical structure and semantic relationship of serial-verb sentences are very complicated,which brings obstacles in its automatic recognition.This paper proposes a recognition model based on neural networks for the recognition of serial-verb sentence.This method uses rules to preprocess the corpus and then applies BERT,the multi-layer CNN and the BiLSTM model to jointly extract features for classification,and then complete the sentence recognition task.Experimental results show that our model achieves an accuracy of 92.71%and F_(1)-value of 87.41%.

关 键 词：连动句 文本分类 神经网络 抽象语义表示 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]