Let's Talk: A Focus on Natural Language and Conversational Dialogue with Bixby
summary
Shows how to train Bixby in Developer Studio using a concrete example capsule, adding annotated NL examples until the IDE marks them learned, compiling the model, and inspecting the planner when an utterance produces an illegal plan. Demonstrates vocab and enum vs name types, contextual continuations that maintain or replace inputs using max one and max many, and implements constraint relaxation and prompt generation so the simulator returns useful results instead of empty lists.
AI-written summary · about
chapters (4)
- 0:00Intro
- 13:53Training Contextual Conversation
- 16:52Introduction to the Conversational Model
- 38:07How will your users speak to your Bixby capsule?
related talks
44mNatural Language Training: Best Practices for Bixby
37mBixby Developer Studio Tour
31mBuilding Bixby Conversational Experiences for Devices with and without Screens
49mBixby: A New Take on the Intelligent Assistant
48mBixby 101: Getting Started
44mGame Changing Voice Assistant: Industry Perspectives
46mBecome a Leader in Voice AI: Join the Bixby Premier Developer Program
01h 04mBuilding Expertise Into Generated Conversations