基于语义理解的叙事文本表示模型  

REPRESENTATION MODEL FOR NARRATIVE TEXTS BASED ON SEMANTIC COMPREHENSION

作　　者：管燕[1] 仲兆满[2] 

机构地区：[1]连云港师范高等专科学校计算机系,江苏连云港222006 [2]上海大学计算机工程与科学学院,上海200072

出　　处：《计算机应用与软件》2009年第7期19-20,30,共3页Computer Applications and Software

基　　金：国家自然科学基金(60575035)

摘　　要：提取文本的特征并用这些特征表示文本是文本信息挖掘的基础工作。揭示了叙事文本是由事件组成的语义特征,给出了基于概念格的叙事文本表示模型。针对事件的某些要素随着时间的更新或者动作的发生而不断改变的客观事实,定义了要素变迁系统来描述这种动态变化过程。The basic work of narrative text information mining is to extract narrative texts＇ feature words and to represent narrative texts by using them. The existing representation models for texts are by means of vector space model or FCA technology, which proceeds from texts＇ feature words and organizes them reasonably to achieve the goal of text representation. However they are lacking of semantic comprehension for texts. Semantic features of narrative texts are focused on in this paper. It reveals the hierachical feathers of narrative texts that they contain a lot of events, and events are composed of event elements. Furthermore, a hierachical representation model for narrative texts based on FCA is proposed.

关 键 词：叙事文本 语义理解 形式概念分析 表示模型 

分 类 号：TP391.7[自动化与计算机技术—计算机应用技术] TP391[自动化与计算机技术—计算机科学与技术]