Learning to read/type a second language in a chatbot enhanced environment

Abstract

Evidence based design methodology can be applied to second language learning by introducing tools and methods based on human machine conversational agents such as restricted chatbots. General purpose chatbots have been used as English tutors, where the learner tries to maintain a generic conversation; on the contrary the proposed tools, obtained by an AIML chatbot generator, are aimed at having a restricted conversation with learners, specifically crafted for the second language training. In the first case study, the obtainable conversation is inspired by the exercises that typically are in a foreign language text book: the chatbot can corrects learners in real time whenever the learner produces incorrect sentences. On the other case, learners have to ask questions to a FAQ-chatbot about a fable that learners should have read to demonstrate the plot of the story has been understood. © 2012 Springer-Verlag.