Bixby Developer Center

With personalization and learning, you can better create capabilities that cater to the habits and routines of users. The models you create for your capsule provide structure to your experience that Bixby can use to learn about its users and the world.

Bixby can learn a lot about users based on their interactions with your capsule. For example, imagine that the user states, "Find me a hotel in Chicago." The user then chooses "Chicago, IL". Bixby learns that the user means "Chicago, IL". The user then adds a follow-up request that states, "Hotels with parking." In this case, Bixby learns that the user has a preference for hotels with parking. Finally, as part of that conversation the user chooses and books "The Drake Hotel / Standard Room, 1 Queen Bed" from a list of results. Bixby learns here that the user likes queen beds when looking for hotel rooms. With each interaction, Bixby is able to automatically learn about the user and your capsule to help them get things done.

This section covers the following topics:

• Selection Learning teaches Bixby how a user makes selections when presented with multiple options in a list.
• Preference Learning ranks and curates the best results from a search action.
• Default Values advises Bixby on how to provide default values.
• Selection Learning Best Practices offers tips on how to write better strategies for selection learning.