基于事件结构的叙事性阅读文本智能分级评测探索

An Exploration of Intelligent Classification Evaluation of Narrative Texts for Reading Based on Event-Structure

叙事性阅读文本筛选对于阅读学习的有效开展有着重要作用。在事件-状态语义模型的基础上,文章从篇章层面出发,对于叙事性文本的事件结构进行了本体和应用的研究,包括定义事件结构,并归纳其特点和类型,形式化描写事件结构,以部编版语文教材中叙事型文本作样本,抽取叙事文本常用事件结构特征,以1047篇叙事性文本作为文本库,分析叙事性文本、事件结构和事件层级等特征,发掘叙事型文本相关特征在阅读活动开展过程中的分布规律,探讨其对叙事性文本分级的影响。

The selection of narrative texts for reading plays an important role in the effectiveness of the reading activity.From the perspective of discourse,the ontology and application of the“event structure”in narrative texts are studied,including defining the event structure and summarizing the characteristics and types based on the Event-State Semantic Model.Then formal description of event structure is researched.The narrative texts in the Chinese textbooks compiled by the Ministry of Education are selected as samples and the event structures of narrative text are extracted to analyze the regularity of its occurrence.1047 narrative texts are used as text library to analyze the characteristics of narrative texts,event structures and event levels to explore the related characteristics of the distribution pattern from the narrative texts in reading activities and the influence on the classification of the narrative texts.