Video Search: Will It Get Better Post Viacom?
April 19, 2013
I know there’s a push to make sense out of Twitter. I know that millions of people post updates to Facebook. I know about text. Searching for text is pretty lousy, but it is trivial compared to video search. Even the remarkable micro-electronics of Glass are child’s play compared to making sense out of digital video flooding the “inner tubes” of the Internet.
This issue is addressed in part in “Why Video Discovery Startups Fail.” Startup video search and discovery systems do face challenges. The broader question is, “Why doesn’t video search work better on well funded services such as Google YouTube or in governmental systems where “finding” a video needle in a digital hay stack is very important?”
The article says:
Video discovery startups are flawed products and even worse businesses. Why? Because they don’t fit into a consumer’s mental model.
The article identifies some challenges. These range from notions I don’t understand like “context” to concepts I partially grasp; namely, monetization.
My list of reasons video search and discovery fails includes:
- The cost of processing large volumes of data
- The lack of software which minimizes false drops
- The time required for humans to review what automated systems do
- The need for humans to cope with problematic videos due to resolution issues
- The financial costs of collection, pre processing, processing, and managing the video flows.
What happens is that eager folks and high rollers believe the hype. Video search and indexing is a problem. If we can’t do text, video remains a problem for the future. Viacom decision or no Viacom decision video search is a reminder that finding information in digitized video is a tough problem which becomes more problematic as the volume of digitized video increases.
Stephen E Arnold, April 19, 2013
Sponsored by Augmentext