Conversational AI 6-DESIGNING AND DEVELOPING DIALOGUE SYSTEMS
There are three main approaches to the design and development of dialogue systems: rule-based, statistical data-driven, and end-to-end neural. In rule-based systems conversation flow and other aspects of the interface are handcrafted using best practice guidelines that have been developed over the past decades by voice user interface designers. These include guidelines on elements of conversations, such as:
- how to design effective prompts
- how to sound natural
- how to act in a cooperative manner
- how to offer help at any time
- how to prevent errors and
- how to recover from errors when they occur.
There are also higher-level guidelines, for example
- how to promote engagement and retention
- how to make the customer experience more personal and more pleasant and
- the use of personas and branding.
Some of these guidelines address linguistic aspects of conversational interaction, such as maintaining the context in multi-turn conversations, asking follow-up questions, maintaining and changing topics, and error recovery. Others are more concerned with social competence, such as promoting engagement, displaying personality, and expressing and interpreting emotion. Finally, there are psychological aspects such as being able to recognize the beliefs and intentions of the other conversational participant. All of these aspects are important for a conversational agent to be effective as well as engaging for the user. In the second and third approaches, dialogue strategies are learned from data.