On diversity of viewing of online video through a large-scale search engine
Yishuai Chen, Changjia Chen, Dan Guo · 2009
We analyze a segment of the user click log with 53.7 m clicks through a large scale video search engine. We find, with the help of search engine, users visit a large number of online videos for a general idea, and visit a lot of Web sites to view a large number of duplications for a popular video content. By studying the relationship between the number of video that a Web site owns and the number of clicks it obtains from the video search engine, we find they are highly correlative, which infers that current video Web sites are highly homogeneous from the view of video search engine. Our observations prove the video search engine keeps the diversity of online video, and it helps users to explore more videos, and compare different copies of same video content to obtain the best result.