This figure shows how the various components of OracleAS Personalization work together. Existing customer data is mapped to the Mining Table Repository schema. The data in this schema is then used by the Mining Object Repository to build packages (mining models). The Mining Object Repository is administered by a user interface to define the Recommendation Engine Farm, packages, and schedule builds, deployments, and reports. The packages are deployed to the Recommendation Engine Farm to generate recommendations, which are requested by the customer applications through the REAPI. The REAPI also enables the applications to collect customer data for the Mining Table Repository.