A Video Retrieval System for Computer Assisted Language Learning

Video has been presented as an effective medium for computer assisted language learning. However, there exist few efficient accessing and managing tools. In this paper, we propose a retrieval system for a large video database. The system retrieves video clips by searching video subtitle text. For language learning, we have designed a syntax search engine embedded in this system. This search engine uses regular expression as the query language and an index construction algorithm is well-designed for speeding up regular expression matching. To ease the burden of authoring lessons from these materials, we implement an automatic video segmentation algorithm to present complete events or actions as final results. The integration of this system and other tools in our authoring environment is also briefly described.