基于科普视频互动弹幕数据挖掘的 受众需求研究——以B 站热门科普视频为例

Research on Audience Demand Based on Data Mining of SciencePopularization Videos Danmaku Texts：Take the SciencePopularization Video on Bilibili as an Example

中文关键词: 科普视频 文本挖掘 弹幕 使用与满足

英文关键词: science popularization video  data mining  danmaku  uses and gratifications

中文摘要:

      以哔哩哔哩视频网站中热门科普视频的弹幕信息为研究对象，利用文本数据挖掘技术，分析受众观 看科普视频的多元化需求，探析优质科普视频交互规律。基于使用与满足理论归纳了受众观看科普视频的四 大需求：科普认知、情感共鸣、社交互动、价值导向，总结了基于弹幕的视频交互特征与规律。最后结合公 众四大需求提出了科普视频设计策略建议：内容翔实、画面直观，讲解语言生动有趣，选取贴近生活的案例， 传播正确科学观念等。

英文摘要:

      Taking the Bilibili danmaku video website as an example，this paper selects the science popularization video and its danmaku texts as the research sample. Data mining and text analysis are used to investigate users’multiple requirements based on the uses and gratifications framework. The results show that the needs of science popularization video watchers consist of scientific knowledge needs，emotional resonance needs，social interaction needs，and value orientation needs. Four main suggestions are given to make compelling science popularization videos：creating detailed content and vivid images；using interesting speaking styles；taking everyday things in life as examples；propagating positive scientific concepts.