基于叙事结构抽象的可视分析方法综述

A Survey on the Visual Analytics of Narrative Structure Abstraction

基于叙事结构抽象的可视分析方法是通过挖掘数据背后存在的叙事逻辑,提供富有洞察力和沟通力的方法.首先介绍叙事结构数据的来源,包括文本数据、图表图像数据、音视频数据、地理空间数据和多模态融合数据;然后讨论叙事结构抽象方法,包括连续性、单元式和离散化3种表示;再探讨叙事结构可视化设计,包括基于视觉通道与表达形式的设计,同时介绍叙事结构交互生成方法,包括基于设计空间和注释的手动生成、AI辅助交互生成和AI自动生成;最后讨论叙事结构可视分析的创作工具以及在文化传媒、医疗健康和体育娱乐等领域的应用,并对该领域现阶段工作的不足进行总结,指出多模态融合数据可视分析、跨平台与多设备的叙事结构展示以及加强情感在叙事结构可视分析中的作用等是未来工作的研究方向.

ing narrative structures through continuous,unit-based,and discrete representations.Additionally,it discusses visual design techniques based on visual channels and visualization genres.The paper presents interactive methods for narrative structures,including manual generation using design space and annotations,AI-assisted interactive generation,and AI-driven automatic generation.Finally,the paper discusses the authoring tools for narrative structure-based visual analytics and their applications in fields like cultural media,healthcare,and sports entertainment.It concludes by summarizing limitations in the field,pointing to future research directions such as visual analytics of multimodal integrated data,cross-platform and multi-device narrative structure presentation,and the enhanced role of emotion in narrative structure visual analytics.