Email
Password
Remember meForgot password?
    Log in with Twitter

article imageToddler robots designed to understand how children to learn

By Tim Sandle     Dec 5, 2016 in Science
Interactive learning of machine could help researchers to understand how children learn new words, according to a new study. Evidence suggests children effectively learn new words in same way as robots.
The surprising connection between the mechanized mind and that of a child arises because of a study that suggests early learning is based not on conscious thought but rather on automatic ability to associate objects. This awareness process allows babies and young children to understand their environment.
According to Dr. Katie Twomey, who is based at Lancaster University, to aim of a recent project was to unpick how young children learn new words for the first time and what they associate these words with.
To explore this, the researchers, which included support from Sussex and Plymouth universities, devised and programmed a humanoid-looking robot which they named iCub.
iCub is a one-meter-high humanoid robot designed for research into human cognition and artificial intelligence. The android was intended to appear similar to a three year old child, in terms of proportions and mass.
The robot was programmed using a software platform that allows the robot to hear words via a microphone and to visualize objects via a camera. The software was a form of artificial intelligence, which allowed the robot to ‘learn.’ The in-built intelligence was sufficient for the robot to be able to point at new objects as it learned the words. This, according to the researchers, is the way that young children ‘learn’ a language.
In her research brief, Dr Twomey explains the learning process: "We know that two-year-old children can work out the meaning of a new word based on words they already know. That is, our toddler can work out that the new word "giraffe" refers to a new toy, when they can also see two others, called "duck" and "rabbit." Here an association is built-up between the shape and color of the toys.
This learning strategy is referred to as "mutual exclusivity", in that learning how a word is associated with an object is based on a process of elimination. In this case, since the brown toy is called "rabbit" and the yellow toy is called "duck," the child works out that the orange toy must be a "giraffe."
The research has been published in the journal Interaction Studies. The research paper is headed “Children's referent selection and word learning.”
More about Language, Learning, Children, Robots
More news from
Latest News
Top News