基于语义依存线索的事件关系识别方法研究.pdf

文档介绍：北京大学学报(自然科学版)
Acta Scientiarum Naturalium Universitatis Pekinensis

基于语义依存线索的事件关系识别方法研究
†
马彬洪宇杨雪蓉姚建民朱巧明
苏州大学江苏省计算机信息处理重点实验室, 苏州 215006; +通信作者, E-mail: tianxianer@
摘要事件关系识别的核心任务是以事件为基本语义单元, 通过分析事件的篇章结构信息及语义特征, 实
现事件语义关系的浅层检测: 即判定任意事件之间是否存在语义关系)。作者通过分析事件的语义依存关系
和事件在演化过程中的语义依存规律, 提出基于语义依存线索的事件关系识别方法。实验结果显示, 相比于
传统的基于语义相似度的识别方法, 新提出的基于事件语义依存线索的事件关系识别方法获得了 5%的性能
提升。
关键词依存线索; 依存分析; 虚拟依存事件; 事件关系
中图分类号 TP391
Using Event Dependency Cue Inference to Recognize Event Relation
MA Bin, HONG Yu†, YANG Xuerong, YAO Jianmin, ZHU Qiaoming
Provincial Key Laboratory puter Information Processing Technology, Soochow University, Suzhou 215006;
† Corresponding author, E-mail: tianxianer@
Abstract Event relation recognization, as one of natural language processing technologies, faces information
stream of texts detecting event relation. The key to event relation recognization is to detect latent semantic relation
(deciding whether events hold logical relation or not) between events by analyzing the corresponding discourse
structure and semantic features of events, with the techniques of semantic relation recognition and inference, which
treats event as the basic semantic unit. This paper proposes an event relation recognization method based on event
dependency cue, by analyzing the semantic dependency relation between events and the rules of eve