基于本体的戏曲类视频语义标注框架构建与应用研究  

Research on the Construction and Application of Semantic Annotation Framework for Traditional Opera Videos Based on Ontology

作　　者：俞娜娜 赵生辉[2] Yu Nana;Zhao Shenghui(School of History and Archives Yunnan University,Kunming Yunnan 650091;School of Management Xizang Minzu University,Xianyang Shaanxi 712082)

机构地区：[1]云南大学历史与档案学院,云南昆明650091 [2]西藏民族大学管理学院,陕西咸阳712082

出　　处：《情报探索》2023年第8期85-93,共9页Information Research

基　　金：西藏高校人文社科基金重点项目“藏学领域多语言知识图谱构建与应用研究”(项目编号:SK2019-03)成果。

摘　　要：[目的/意义]针对戏曲类视频资源内容的管理难题,旨在构建一个戏曲类视频领域本体,以情节、事件、实体、动作、戏曲音乐等为核心要素,为戏曲类视频语义标注提供参考。[方法/过程]从分析戏曲类视频内容特征出发,确定本体应覆盖的范围;选择复用与参考已有的相关本体,在此基础上进行扩展,确定核心要素及其之间的关系,使用Protégé构建本体。[结果/结论]以豫剧和昆曲的两个经典剧目为例进行语义标注,通过SPARQL检索说明了标注框架的有效性。该本体可对戏曲类视频内容语义有效描述,将多方面的戏曲领域知识与视频内容进行关联,实现视频内容的结构化表达,为向不同类型的用户提供戏曲类视频内容检索提供关键基础。[Purpose/significance]According to the management problems of thetraditional opera video resources content the paper aims to build an ontology of traditional opera video domain with plot event entity action traditional opera music etc.as the core elements to provide reference for the semantic annotation of traditional opera video.[Method/process]Based on the characteristics of traditional opera video content the paper determines the scope that the ontology should cover selects reuse and refers to existing related ontology expand on this basis determine the relationships among core elements and uses Protégéto build ontology.[Result/conclusion]It takes two classical operas of Henan Opera and Kunqu Opera as examples for semantic annotation.The validity of the annotation framework is illustrated by SPARQL retrieval.This ontology can effectively describe the semantics of traditional opera video content associate various aspects of traditional opera domain knowledge with video content and achieve structured expression of video content and provide a key foundation for traditional opera video content retrieval for different types of users.

关 键 词：视频标注 语义标注 事件本体 领域本体 

分 类 号：G254[文化科学—图书馆学]