教学视频信息文本化检索平台的构建与应用

发布时间：2018-05-12 23:22

本文选题：教学视频 + 搜索引擎　；参考：《华东师范大学》2012年硕士论文

【摘要】：随着Internet的迅速发展和教育信息化的推广,教学视频资源在教育界得到了广泛的关注和使用。就我国目前的教育领域的现状来看,国家,学校,公司或者一些教育网站都建立了有关教学视频的资源库,但是现有的教学视频资源在易检索性、实用性、交互性等方面存在着问题。面对海量的教学视频资源,如何克服存在的问题,让用户能够随时、快速、较准确地检索到目标资源,日趋显得重要。本文首先分析了目前教学视频资源的现状与存在的问题,对目前主流的基于内容的视频检索技术、相关视频检索系统,以及Lucene检索引擎的相关研究现状进行了分析；梳理和归纳了搜索引擎的工作机制、全文检索(系统)、分词技术、排序技术、评价指标,以及Lucene的主要技术特点。之后对教学视频检索系统的需求、总体结构和核心模块设计进行了深入的研究和分析；对系统平台中集成的工具集的设计及实现进行了阐述；最后根据检索系统的需求和设计思路,基于检索引擎Lucene,设计并实现了一个高性能的教学视频信息文本化检索系统。系统采用了数据库检索、全文检索以及语义检索整合的混合检索方式,互相弥补自身的不足之处；此外依托平台实现教学视频信息文本化检索平台套件工具集的设计与开发,使得系统平台可以作为集教学视频标注、评价、视频共享和视频检索的一体化环境。本文还对检索系统进行了测试和分析。测试数据和分析结果表明,教学视频检索系统的检索效率高,查全率、查准率以及用户满意度都达到了设计的要求；最后对系统平台的应用场景及应用模式进行探索与实践。
[Abstract]:With the rapid development of Internet and the promotion of educational informatization, instructional video resources have been widely concerned and used in the field of education. As far as the current situation in the field of education in China is concerned, the country, schools, companies or some educational websites have all established a resource bank for teaching videos, but the existing teaching video resources are easily searchable and practical. There are problems in such aspects as interactivity. In the face of massive teaching video resources, how to overcome the existing problems and enable users to retrieve target resources at any time, quickly and accurately, is becoming increasingly important. This paper first analyzes the current situation and existing problems of instructional video resources, and analyzes the current mainstream content-based video retrieval technology, related video retrieval systems, and the research status of Lucene retrieval engine. This paper summarizes the working mechanism of search engine, full-text retrieval (systematical search, participle technology, sorting technique, evaluation index, and main technical features of Lucene). After that, the requirements, the overall structure and the core module design of the instructional video retrieval system are deeply studied and analyzed, and the design and implementation of the integrated toolset in the system platform are described. Finally, according to the requirements and design ideas of the retrieval system, a high performance text-based retrieval system for teaching video information is designed and implemented based on the retrieval engine Lucene. The system adopts the mixed retrieval methods of database retrieval, full-text retrieval and semantic retrieval to make up for their shortcomings. In addition, the design and development of the toolset of text-based retrieval platform for teaching video information is realized based on the platform. The system platform can be used as an integrated environment for teaching video annotation, evaluation, video sharing and video retrieval. This paper also tests and analyzes the retrieval system. The test data and analysis results show that the retrieval efficiency, recall rate, recall rate and user satisfaction of the instructional video retrieval system meet the requirements of the design. Finally, the application scene and application mode of the system platform are explored and practiced.
【学位授予单位】：华东师范大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TP391.3

【参考文献】