Relation Classification: CNN or RNN?

摘要：

　　Convolutional neural networks(CNN)have delivered competitive performance on relation classification,without tedious feature engineering.A particular shortcoming of CNN,however,is that it is less powerful in modeling longspan relations.This paper presents a model based on recurrent neural networks(RNN)and compares the capabilities of CNN and RNN on the relation classification task.We conducted a thorough comparative study on two databases: one is the popular SemEval-2010 Task 8 dataset,and the other is the KBP37 dataset we designed based on MIML-RE 1,with the goal of learning and testing complex relations.The experimental results strongly indicate that even with a simple RNN structure,the model can deliver much better performance than CNN,particularly for long-span relations.

作者: Dongxu Zhang Dong Wang

作者单位: CSLT,RIIT,Tsinghua University;PRIS,Beijing University of Posts and Telecommunications Beijing,China CSLT,RIIT,Tsinghua University;Tsinghua National Lab for Information Science and Technology

会议类型: 国际会议

会议名称: 第五届自然语言处理与中文计算会议(NLPCC-ICCPOL2016)

会议地点: 昆明

会议语种:英文

页码: 1-10

在线出版日期: 2016-12-02（万方平台首次上网日期，不代表论文的发表时间）