| With the emergence of Web2.0, the video websites has been developed rapidly. The Video has become an important way to get information, but the large video sites video on the growing number of choices to the user at the same time, forcing users to spend a lot of time to check the video they want, so the video recommended Become the critical needs of large-scale video of site. Video semantic annotation through video content embodied in their assigned numbers according to semantic concepts, in this video clip can be achieved on the basis of recommendations.Video annotation on the current work to make a study found that whether it is pure hand-tagging or auto-tagging based on machine learning, there are some problems and difficulties, so we use a combination of both semi-automatic annotation methods, namely the use of video Web Web2.0 the characteristics of the Aspect collection of video content the user described on the basis of this description were accurate video processing fine-grained annotations. Video of education as the research object, explore the use of covert video video users to mark the way to learn the user's use of video to share the process of fine-grained access to video annotation, so that users can combine the distribution of interest and obtain the recommended time point marked fragments and labeling concepts for video recommendation of great value.Video Aspect information with self-learning system, students can take advantage of the characteristics of independent learning by supporting the user in ways that students rely on the video network and increase user stickiness, increasing opportunities for users to make video description. Finally, in the system implementation process, several key technologies - AJAX technology, memcached cache, flash and other technologies play a more detailed study, the last use of php web development language such a mature system development completed. |
