Uh oh! We couldn’t find any match.
Please try other search keywords.
Instantiation Strategies inform Bixby's execution about where to get missing values that are required in order to fulfill a user request. During execution, an action input may require one or more values that the user request or executing context has not specified. Each matching
instantiation-strategy is evaluated and can provide one or more values to planning. If after all strategies have been evaluated, more than one instantiated option exists, it is passed to Selection Learning for the system to decide which one, if any, should be automatically selected for the user.
There is no learning for Instantiation Strategies. If they match, are valid, and provide results, they are passed on to execution, where they are merged and used in other Bixby learning subsystems for processing.
If no instantiation strategy is provided, or the provided strategies do not resolve to any values, the user is prompted for the missing information.
The instantiation strategy below provides a set of possible tip percentage values when required. For example, if calculating a total bill requires a tip percentage and the user hasn't specified one in the request this, along with other matching strategies, provide execution with possible tip percentages for the request:
Since this strategy's
intent specifies more than one tip percentage, it is guaranteed to invoke Bixby's Selection Learning to help pick one automatically.
|Optional description of the strategy for documentation purposes|
|Identifier that is unique for a given capsule|
|Allows the developer to define the specific scenarios that a Instantiation Strategy should apply|
|A strategy block encompasses the intent that sources the instantiated value(s)|