A Multi-Layer Architecture for Intelligent Chatbots in IoT Robots: Design and Implementation

This thesis designs and implements an intelligent chatbot system called Rima, designed for IoT robots. The system is based on a three-layer architecture that integrates language understanding services (LUIS), a retrieval-based chatbot (AIML), and a generative chatbot (Seq2seq), enabling command recognition and natural language conversation. Through cross-platform development, Rima operates in various client environments and offers API support, allowing the chatbot to be utilized in IoT devices. The experiments demonstrate that the multi-layer hybrid structure performs well in both command execution and casual conversations, proving to be highly practical.

Thesis