Conversational Agents and Dialogue Systems
Software agents that carry on natural-language conversations with people, spanning task-oriented assistants and open-domain chatbots. Covers the core components of a dialogue system (understanding user input, tracking conversational state and context, deciding what to say, and generating a response), the distinction between reactive agents that reply only when addressed and agents that manage multi-party interaction, and how modern systems are built on large language models. Also covers the challenge of deciding not just what to say but when to speak, especially when several participants share the floor.
Spoken Dialog Systems Task Oriented Dialogue Architecture and Design Principles
Task-oriented dialogue architectures operate on the principle that effective human-machine interaction requires a dedicated natural language understanding (NLU) and dialog management stage beyond sim…