SOTA is a robot developed by NTT Data with the aim to provide an assistant or a companion. Our team customised it by giving it the ability to tell interactive stories, suited for young children. The idea behind the project originates in a piece of research that shows that there is a considerable “word gap” between children coming from different backgrounds, children coming from high-income families hearing up to 30 million more words than children from lower class. Given the plasticity of the brain in the early stages of development, it is crucial for kids to hear as many words and ideas as possible. This project is to be used in further research as a tool for gathering data, in order to gain more insight into how children acquire a language and develop social skills.
By using IBM Watson Conversation, we were able to extract meaning from the user’s speech and match it to available path choices in the stories
We convert user speech into text for later processing using Google Speech API
With the help of IBM Watson, we were able to create speech which is strikingly similar to that of humans, using voice inflexions and tonality changes
A carefully chosen set of gestures ensures that SOTA feels friendlier and more personified
Upon not knowing a word, a child can simply ask SOTA for its meaning
For the purpose of research, all the user replies are stored in a MySQL database for later analysis
