San Francisco - A former video store owner on Tuesday launched ClerkDogs, a San Francisco-based movie recommendation service that mixes technology with the movie analysis of actual former video store clerks. Founder Stuart Skorman, one of the pioneers of the recommendation technology known as collaborative filtering used by Netflix and other brands, believes that technology has taken the movie prediction business as far as it can go, and "requires a mix of man and machine similar to complementary destinations such as Wikipedia for information or Pandora for music." ClerkDogs' core tool uses 35 mood-defining attributes that can be adjusted, giving users a list of movie matches selected by the clerks. The company is now in beta mode, featuring films specifically in the crime and suspense genres, with other genres to be added. Skorman opened his first brick and mortar video store in 1985 and moved to the web in 1995 when he founded Reel.com, which he sold three years later to Hollywood Entertainment for $100 million.