【摘要】 目前教学网站提供的教师讲课视频一般以课时为单位,时长很长,而且知识点繁多。本文提出了一种可以快速将视频根据知识点进行剪辑整理的方法。该方法首先采用语音识别技术将视频中的语音转化成文字,生成字幕文件,然后借助字幕文件中关键词出现的时间段对整个教学视频进行分节或提取知识点,制作成带有字幕的教学短视频。采用这种根据知识点分节的短视频,学生可以依据自己的兴趣重点搜索、学习相关知识点视频,而不是整个视频,从而大量减轻学生负担,提高学习效率。更多还原

【Abstract】 Presently the videos on most teaching websites are provided in time unit of the class hour, which are relatively long and consist of lots of knowledge points. The paper proposed a method to clip the video quickly according to knowledge points. The method first uses the speech recognition technology to convert the speech in a teaching video into text, producing a subtitle file. Then, it divides the whole video into sections or extracts the knowledge points based on the time stamp of the keywords in the subtitle file,yielding short teaching videos with subtitles. With such short videos students may search and focus on their interested knowledge points without referring the whole video.更多还原