Adding predictors to existing adaptive models
For new implementations of Pega Customer Decision Hub, when you save the Context Dictionary configuration, the journey predictors are automatically added to the adaptive models. If you are upgrading your application from version 8.5 or earlier, the predictors are not added automatically, and you must add them manually.
For more information about Context Dictionary, check Defining customer contexts for multilevel decisioning in Pega Customer Decision Hub.
Pega Customer Decision Hub
You must add the predictors to the strategy models at two different levels: action and treatment.
- Log in as an operator with access to Dev Studio.
- In the navigation pane of Dev Studio, click
- Select the strategy to which you want to add predictors
- To add predictors to an action-level strategy model, select ActionModelImpl. The ruleset of the model should be Artifacts.
- To add predictors to a treatment strategy model, select the treatments strategy, for example, WebTreatmentModelImpl. Every channel has a different treatment model.
- Right-click on Adaptive model component and choose Open Adaptive model.
- On the Predictors tab, click .
- Add the following parameters:
- Make sure that the Data type for all added parameters is Text, and that Predictor type is set to Symbolic, as in the following figure:
- Click Save, and go back to your strategy.
- Right-click the Adaptive model component on the strategy canvas, and then click Properties.
- In the Parametrized predictors section, map values to
See the following figure for reference:
- To the Journey parameter map .pyJourney value.
- To the JourneyStage parameter map .pyStage value.
- To the Last Journey Stage parameter map .LastStage value.
- To the DaysinCurrentStage parameter map .DaysInCurrentStage value.
- To the PriorStageInJourney parameter map .PriorStageInJourney value.
Previous topic Understanding how customer journeys influence Next-Best-Actions Next topic Filtering out actions not associated with the current customer journey stage