【作者】 钟茂生；

【Author】 Zhong Maosheng~(1,2) 2.School of Information Engineering,East China Jiaotong University,NanChang,330013;2.Department of Compute Science and Engineering,Shanghai Jiaotong University,ShangHai 200240

【机构】 华东交通大学信息工程学院； 上海交通大学计算机科学与工程系；

【摘要】 句际语义关系是指语篇上下文中相邻句子之间存在的语义关系。准确的识别句际语义关系对于文本理解、文本推理和文本结构分析具有重要的意义。然而,由于受到语篇上下文环境、指代消解、句法结构等多个因素的影响,使得句际语义关系自动分析研究具有很大的困难。该文在语言学界总结的句际语义关系和句际语义关系对应的词语形式标记的基础上,提出了一种机器自动识别上下文句际语义关系的方法,包括词语形式模板的获取、模板冲突消解的方法以及句际语义关系识别算法。实验结果表明,该文提出的方法具有较好的识别效果。更多还原

【Abstract】 The inter-sentence semantic relation is a semantic relation where exists between adjacent sentences in the context of the discourse.Accurate recognizing of inter-sentence semantic relation is of great significance to text understanding,text reasoning and text structure analysis.However,because of the impact of a number of factors,such as discourse context environment,anaphora resolution and syntactic structure,these make that the research on automatic recognizing the inter-sentence semantic relation is very difficult.On the base of the classification of inter-sentence semantic relation and the words sign template corresponding to a inter-sentence semantic relation summarized by the language researcher,this paper presented a method on automatic recognizing the inter-sentence semantic relation in the context of the discourse,included the way of obtaining the words sign template,the way of template conflict resolution and the algorithm of recognizing the inter-sentence semantic relation.The experiment results show that the proposed method has high performance for recognizing the inter-sentence semantic relation.更多还原