The study “Make Your Museum Talk: Natural Language Interfaces For Cultural Institutions” was done in 2002 by Stefania Boia, Giuliano Gaia and Morgana Caldarini.
The chatbot was made to answer questions about Leonardo Da Vinci and his life on the website of the Science Museum.
Instead of trying to recreate Leonardo himself, the chatbot resembled one of Leonardo’s machines. “We decided to avoid the obvious choice of making it look like Leonardo, because it would have been banal and very difficult to realize – how can you in fact program software to talk like a genius?”
Making it one of Leonardo’s machines also had other advantages, including being able to justify some of the bots errors and having the freedom to give the character more gestures and facial expressions.
Unfortunately, the Leonardo project was stopped before it could be thoroughly tested and launched.
However, the initial tests showed good results and positive user feedback. The creators had big plans for their bot, including SMS-features, making him talk, using voice input and creating interactions with other museum bots.