Question and Answer Generation from Messages or Voice Announcement using Similarity and One Learning Neuron

摘要

This paper proposes a question and answer (Q&A)-type operation interface (QATOI) between an elevator and passengers in a double-deck elevator system that can utilize messages or a voice announcement-type operation interface (VATOI). In our proposed structure, a one learning neuron (OLENE) system with a teacher in each layer then selects an appropriate answer or response to the question/message by utilizing a back-propagation method based on simple sentence similarity (SSS).