Categorizing Big Video Data on the Web: Challenges and Opportunities

Video categorization is a very important problem with many applications like content search and organization, smart content-aware advertising, open-source intelligence analysis, etc. This paper discusses selected representative research progresses in categorizing big video data, with a focus on the user-generated videos on the Internet. We identify two major challenges in this vibrant field and envision promising directions that deserve in-depth future investigations. The discussions in this paper are brief but hopefully useful for quickly understanding the current progress and knowing where we should go in the next couple of years.