数据故事中的人物类型及自动生成方法研究  

Study on Character Type and Automatic Generation Method in Data Story

作　　者：朝乐门[1,2] 刘慧 张天怡 李泽仑 Chao Lemen;Liu Hui;Zhang Tianyi;Li Zelun(Key Laboratory of Data Engineering and Knowledge Engineering(Renmin University of China),Beijing 100872;School of Information Resource Management,Renmin University of China,Beijing 100872)

机构地区：[1]数据工程与知识工程教育部重点实验室(中国人民大学),北京100872 [2]中国人民大学信息资源管理学院,北京100872

出　　处：《图书情报工作》2023年第24期99-110,共12页Library and Information Service

基　　金：国家自然科学基金项目“预测性分析结果的数据故事化描述方法及关键技术”(项目编号:72074214)研究成果之一。

摘　　要：[目的/意义]人物和情节是数据故事的两大支柱。数据故事的情节通过人物特征、行为、所期待目标、所面对现实和所认为偏见来展开,实现数据故事人物的自动化生成是数据故事化领域科学研究的核心主题之一,对于数据故事的理论研究、自动生成和工程化研发具有重要意义。[方法/过程]首先,探讨数据故事人物的类型、特征及操作。其次,提出基于反事实解释的人物生成方法,分别对数据故事中的主人公、同类人物、异类人物、正面人物和反面人物给出自动生成方法。接着,分析其技术实现,探讨实验设计、数据来源、方法选择及结果讨论。最后,总结论文的主要研究发现,并对未来研究提出建议。[结果/结论 ]在数据故事化领域首次较为系统研究数据故事人物的组成要素、基本类型、主要特征及核心操作,并提出基于反事实的数据故事人物自动生成方法。[Purpose/Significance] Characters and narratives form the dual pillars of data stories. The narrative trajectory of a data story is shaped by the characters' traits, behaviors, goals, realities, and biases. Achieving automatic generation of data story characters stands as a core scientific research topic in the realm of data storytelling.Addressing this central issue carries significance for the theoretical exploration, automation, and engineering-oriented research and development in the domain of data stories. [Method/Process] Initially, this study delved into the types, attributes, and operations associated with data story characters. Subsequently, it proposed a character generation technique based on counterfactual reasoning, offering automatic generation algorithms for protagonists,similar characters, heterogeneous characters, positive characters, and negative characters in data stories. Following this, it dissected and discussed its technical implementation, furnishing the paper's experimental design, data sources, method selection, and result discussion. Lastly, it encapsulated the principal research findings of the paper and furnished a forward-looking perspective. [Result/Conclusion] In the data storytelling, this paper introduces for the first time the compositional elements, basic types, principal features, and core operations of data story characters.Furthermore, this study presents an automatic generation method for data story characters rooted in counterfactual reasoning.

关 键 词：数据故事 数据故事化 故事人物 人物 

分 类 号：G203[文化科学—传播学]