Many attempts to make computers more personable have failed - like the famously irritating animated paperclip, otherwise known as Microsoft's "office assistant". But perhaps we would enjoy interacting with our computers more if they mimicked our speech patterns. At least that's what Noriko Suzuki's team at the ATR Media Information Science Laboratories in Kyoto thought. They got the idea from research on how social bonds develop between people. For instance, we develop a rapport with a child when it mimics us; we even do it with a parrot that is merely, well, parroting us. Knowing when someone is trying to build up a rapport with you is key to good social relations, they say.
Timothy Bickmore, an expert on human-computer interaction at the Massachusetts Institute of Technology Media Lab in Cambridge, agrees. "Mimicry and mirroring between two people is correlated with increased solidarity and rapport," he says. "It's something that therapists do to build solidarity with their clients."
To see if computers could establish such a rapport with their users, Suzuki asked some volunteers to work on screen with an animated character that they were told had the speech skills of a one-year-old child. Their task was to make toy animals out of building blocks on the screen, and at the same time teach the character the names of the toys being built.
In response, the character hummed back sounds that mimicked characteristic features, such as the rhythm, intonation, loudness and pitch of the user's voice. The extent of the mimicry varied.
The users then rated the character in areas such as cooperation, learning ability, task-achievement, comfort, friendliness, and sympathy. The animated character scored highest on all these factors when its voice was mimicking about 80 per cent of the user's voice. The user felt some kind of friendly emotion from the computer, even though it was just copying the stresses and intonation of their own voice, says Suzuki. The researchers think that the 20 per cent of sound that is not mimicked gives the user the impression that the character has some degree of free will.
Suzuki thinks that embedding such mimicry into speech during vocal interactions between computers or robots and humans will make things better. "Sometimes people are afraid of robots," says Suzuki. But if robot voice patterns are improved, people may warm to them, he adds. MIT's Bickmore agrees that such mimicry could help people build a rapport with computer characters and robots, and sees applications in entertainment, computer gaming, and toys.
Author: Anil Ananthaswamy
New Scientist issue: 31 MAY 2003
UK CONTACT - Claire Bowles, New Scientist Press Office, London:
Tel: 44-207-331-2751 or email claire.bowles@rbi.co.uk
US CONTACT - Michelle Soucy, New Scientist Boston Office:
Tel: 617-558-4939 or email michelle.soucy@newscientist.com
PLEASE MENTION NEW SCIENTIST AS THE SOURCE OF THIS STORY AND, IF PUBLISHING ONLINE, PLEASE CARRY A HYPERLINK TO: http://www.newscientist.com
"These articles are posted on this site to give advance access to other authorised media who may wish to quote extracts as part of fair dealing with this copyrighted material. Full attribution is required, and if publishing online a link to www.newscientist.com is also required. Advance permission is required before any and every reproduction of each article in full - please contact celia.thomas@rbi.co.uk. Please note that all material is copyright of Reed Business Information Limited and we reserve the right to take such action as we consider appropriate to protect such copyright."