Proposal of Learning Support System for Improving Skills in Inferring Background Knowledge in Conversation

Abstract

In communication, a claim should be conveyed to others correctly. However, establishing such communication is not easy. In this study, we define miscommunication as a situation where a claim is conveyed incorrectly. We discuss how to resolve miscommunication and propose a learning support system designed to help resolve

In considering how to resolve miscommunication, we constructed a dialogue model consisting of three elements: claim, background knowledge, and utterance. The claim is an individual’s internal demands, the background knowledge is information for constructing claims, and the utterance is the text for conveying claims to others. In dialogue, speakers convey claims to listeners. In conveying a claim, the speaker constructs an utterance(s) based on their background knowledge. The listener receives the utterance(s) and infers the intended meaning of the speaker’s claim based on their own background knowledge. Therefore, the difference in the claim conveyed and the claim received is a difference in background knowledge or a difference in interpretation of the utterance or error of inferring.

In this paper, we focus on the difference in background knowledge and propose a learning support system for improving skills in inferring others’ background knowledge. This system involves an example of miscommunication between a speaker and a listener, where learners constructs the listener’s background knowledge.