Pavideoge: A Metadata Markup Video Structure in Video Search Engine

In this paper, we study the problems of video processing in video search engine. Video has now become a very important kind of data in Internet; while searching for video is still a challenging task due to the inner properties of video: requiring enormous storage space, being independent, expressing information hiddenly. To handle the properties of video more effectively, in this paper, we propose a new video processing method in video search engine. In detail, the core of the new video processing method is creating pavideoge--a new data type, which contains the video advantages and webpage advantages. The pavideoge has four attributes: real link, videorank, text information and playnum. Each of them combines video's properties with webpage's. Video search engine based on the pavideoge can retrieve video more effectively. The experiment results show the encouraging performance of our approach. Based on the pavideoge, our video search engine can retrieve more precise videos in comparsion with previous related work.