摘要
国外一批计算语言学研究者为开发语篇结构自动生成系统,以修辞结构理论为支撑,对大量英语语篇进行标注后发现,语篇中的修辞关系应加以限制,语篇的抽象结构应为树型图式。这些国外研究成果带动了汉语篇章结构研究的发展,汉语语篇也应由基本语篇单位构成,其抽象结构也应为树型图,但基本语篇单位的切分并不是以引导从句的连词或短语为标记而是以逗号为切分点。在对语篇中的修辞关系进行分析时,可套用修辞结构理论的关系集。
In order to develop automatic summarization systems, a group of computational linguists annotated a large number of English texts on the basis of Rhetorical Structure Theory, from which they inferred that rhetorical relations between elementary discourse units should be restricted and the abstract structures of texts should be tree-type graphs. This result has brought profound implications for the research on Chinese text structures. Based on the research done by foreigners, it can be concluded that Chinese texts consist of elementary discourse units and their abstract structures should be tree-type graphs. However, what is different is that elementary units in Chinese texts are signed by comma instead of subordinate clauses beginning with functional words or phrases. In the course of analysis of Chinese texts, Rhetorical Structure Theory can be referred to or even applied directly.
出处
《长春大学学报》
2011年第1期39-43,共5页
Journal of Changchun University
基金
上海市哲学社会科学规划课题研究基金项目(2006EYY002)
关键词
修辞结构理论
基本语篇单位
修辞关系
汉语篇章结构
Rhetorical Structure Theory
elementary discourse unit
rhetorical relation
Chinese text structure
