The syntactic and semantic analysis of interrogative sentences has a wide application in the fields of search engines,information extraction and question answering systems.The NLP systems usually use a combination of classification and syntactic analysis to process interrogative sentences,with poor accuracy and efficiency.The interrogative sentence has rich linguistic research results,such as interrogative sentence structure types,etc.,but it lacks systematic formal representation.We use Chinese Abstract Semantic Representation(CAMR)based on graph structure to annotate.The data comes from Penn Chinese Treebank 8.0,Chinese textbooks for elementary schools,and the Chinese translation of Little Prince,for a total of 2071 sentences.All kinds of interrogative words are represented by the combination of the interrogative concept-amr-unknown and the semantic relationship,which can represent the key information of the interrogative sentence,the question focus and the semantic structure of the interrogative sentence.Finally,we calculate the probability distribution of the focus,of which the cause,modifier,and argument accounted for the highest proportion,respectively accounting for 26.53%,16.73%,and 16.44%.Interrogative sentences annotating and analysis based on abstract semantic representation provides a better theory and resources for the study of Chinese interrogative sentences.
疑问句抽象语义表示语义角色中文信息处理
闫培艺、李斌、黄彤、霍凯蕊、陈瑾、曲维光
展开 >
南京师范大学文学院,江苏南京
南京师范大学计算机科学与技术学院,江苏南京
疑问句 抽象语义表示 语义角色 中文信息处理
Chinese National Conference on Computational Linguistic
Haikou(CN)
19th Chinese National Conference on Computational Linguistic
NETL
