Filtering and Ranking
How the building blocks are used to produce recommendations The purpose of filtering and ranking offers is at the heart of the Best-Next-Action Decisioning process. The process can be summarized as follows: fn getRecommendations(Offers, Context, Options) -> list[Recommendations] Offers and context is used to produce a ranked set of recommendations for the customer. The steps in the filtering process are as follows: Start with the pool of offers that have been constructed Remove all of the offers that do not support the channel requested in the context Remove all of the offers that are not currently valid based on the current date and time Remove all of the offers whose set of inclusion rules do not evaluate to True Remove all of the offers for which any of the exclusion rules evaluate to True The steps in the ranking process are as follows: