【作者】 陈雪艳； 吕国英； 李茹； 刘伟；

【Author】 Chen Xueyan Lu Guoying Li Ru Liu Wei School of Computer & Information Technology,Shanxi University,Taiyuan,Shanxi 030006,China

【机构】 山西大学计算机与信息技术学院；

【摘要】 本文提出了一种基于层叠条件随机场的CFN句法语义自动标注方法。该方法在低层条件随机场模型中解决了框架元素的识别,将识别结果传递到上层短语类型识别的条件随场模型,再将识别结果传递到上层句法功能识别的条件随机场模型,其低层模型为上层模型提供决策支持,并且在每层自动标注完成后,增加后处理规则去识别那些没有被正确标注的词语。实验选用CFN中"陈述"框架下的句子库,实现了基于层叠条件随机场句法语义自动标注的原型系统。更多还原

【Abstract】 This paper presents an approach of Syntactic and Semantic of CFN automatic tagging based on cascaded conditional random fields model.The frame elements are first recognized by the lower random fields model .The result then is passed to the upper phrase type identification with the random fields mode,and then transfer the results to identify the upper syntactic function identification with the random fields mode.The lower model supports the decision for high model and adds post-processing rules for the labeling blocks which is unrecognized alter each automatic hierarchical tagging.The experiment selects the sentence of"Statement"frame in CFN,has realized Syntactic and Semantic automatic tagging based on cascaded conditional random fields model.更多还原