摘要
针对交互式的多媒体学习系统的特点,提出了一种基于自然语言的方法来实现基于内容的视频检索,用户可以用自然语言和系统进行交互,从而方便快捷地找到自己想要的视频片段。该方法集成了自然语言处理、实体名提取、基于帧的索引以及信息检索等技术,从而使系统能够处理用户提出的自然语言问题,根据问题构建简洁明了的问题模板,用问题模板与系统中已建的描述视频的模板进行匹配,从而降低了视频检索问题的复杂度,提高了系统的易用性。
A new approach to video indexing and retrieval is presented according to the peculiarity of online learning system.The natural language in the system is used to get the video clip needed expediently.The method integrates natural language processing,named entity extraction,frame-based indexing,and information retrieval techniques to make the system process questions proposed by natural language, question template is created according to the question,then the system try to find the best match between the question template and the templates created based on videos,it reduce the complexity of video retrieval and make the online learning system easier to use.
出处
《计算机工程与设计》
CSCD
北大核心
2010年第3期647-650,共4页
Computer Engineering and Design
关键词
在线学习
自动问答系统
自然语言处理
问题模板
视频索引与检索
online learning
learning by ask
natural language processing
question template
video indexing and retrieval
