Wales Wikia Search Beached
March 31, 2009
In London at the International Online Meeting a couple of years ago, I listened to a talk by Jimmy Wales. He was explaining the idea of different type of search system. What he referenced emerged as Wikia Search. Bouyed by the success of Wikipedia (the online encyclopedia that slew Encarta), Mr. Wales rolled out a user-generated serach engine. My view of the idea remained neutral. When humans are involved in search, the problems come out of the wood work. Say what yoiu want about Google and its math recipes, but at least there are reasonably objective. When humans pick sites and index, objectivity becomes a matter for discussion. According to the CNet story here, usage was modest, about 10,000 uniques per month.
User generated search listings are a great idea in theory but tough to operate over time an ensure consistency. Other challenges of user generated services include:
- Humans want to be recognized with money or fame. Algorithms work without too much threat of unionization, slow downs, or quitting
- Humans get bored. Algorithms don’t, at least not yet
- Humans volunteer and then only a couple of individuals carry the freight. Algorithms can be spawned and killed when they are no longer needed.
Our 1993 Point service relied in humans and algorithms. We sold it. I try to stick with automation. Some clever folks will find a way to make human labor pay off in the search and content processing space. These will be niche plays because of cost and speed considerations (more accurately, throughput).
Now that Wikia search has reached the beach to die, which human-powered search system is next?
Stephen Arnold, March 31, 2009