情报科学 ›› 2023, Vol. 41 ›› Issue (8): 104-112.

• 业务研究 • 上一篇    下一篇

弹幕视频网站中用户社会交互行为特征分析
——以B站美食制作类视频为例

  

  • 出版日期:2023-08-01 发布日期:2023-09-19

  • Online:2023-08-01 Published:2023-09-19

摘要:

【 目的/意义】探究弹幕用户社会交互行为特征,一方面帮助视频网站了解弹幕用户行为特点,重视弹幕的 过滤引导;另一方面帮助美食制作类视频创作者了解用户交互行为倾向,合理设置情境,提高视频内容质量。【方 法/过程】选取B站美食区中30个美食制作视频,统计分析用户弹幕交互行为在时间、数量以及文本方面的特征,运 用情感分析和主题词分析探究用户互动内容特点,通过社会网络图分析用户间的跟随行为。【结果/结论】弹幕用户 倾向午饭、晚饭以及凌晨时间段进行社会交互,弹幕峰值随交互情境的发生而形成。弹幕用户情感倾向与视频创 作者内容表述和情绪传达有关。弹幕用户与视频创作者互动多用“我”或“我们”为主语或使用创作者ID交流,弹幕 用户之间倾向寻找具有共同点或相似经历的人。“活跃用户”互动内容具有话题融合度高的特点,跟随者的二次创 作特征明显。【创新/局限】文章着重关注弹幕用户的社会交互行为,多个维度探究行为特征,创新性将弹幕文本分 析与美食制作类视频弹幕相融合;后续研究可进一步扩大样本量和细化视频分类。

Abstract: 【 Purpose/significance】 To explore the characteristics of social interaction behavior of bullet screen users, on one hand, to help video websites understand the characteristics of bullet screen users' behavior, attach importance to the filtering and guidance of bullet screen; On the other hand, it helps creators of food production videos to understand the user's interactive behavior tendency, set the situation reasonably, and improve the quality of video content【. Method/process】 30 food production videos were selected from the food area of Station B, and the features of user interaction in terms of time, quantity and text were statistically analyzed. Emotion analy⁃ sis and subject word analysis were used to explore the features of user interaction content, and the following behaviors among users were analyzed through social network graph.【 Result/conclusion】 Bullet screen users tend to have social interaction during lunch, din⁃ ner and early morning, and the peak value of bullet screen is formed with the occurrence of interactive situations. The emotional ten⁃ dency of bullet screen users is related to the content expression and emotional transmission of video creators. When interacting with video creators, bullet users often use "I" or "we" as the main language or communicate with creator ID. Bullet users tend to look for people with common ground or similar experiences. The interactive content of "active users" has the characteristics of high degree of topic integration, and the characteristics of secondary creation of followers are obvious.【 Innovation/limitation】 This paper focuses on the social interaction behaviors of bullet screen users, explores their behavioral characteristics from multiple dimensions, and innova⁃ tively integrates bullet screen text analysis with food production video. Follow-up studies can further expand the sample size and re⁃ fine the video classification.