Research and Application of Event Finding based on massive Internet Imagine Tag

System mainly studies mass events found images from the Internet, this paper focuses on the data label document Flickr to quantify. This paper also implements single-pass clustering algorithm using traditional text clustering. In this paper, achieve three strategies of single-pass clustering algorithm, and analyze and compare the three strategies. Different document order in the event of the discovery process requires traditional events found single-pass clustering algorithm. This paper implements a hot event with a buffer to find a way.