Predicted response design

The Predicted response is also of the JSON format. A simple example for our example of the result is the following JSON:

    {
      "itemScores":[
         {"item":22,"score":4.07},
         {"item":62,"score":4.05},
         {"item":75,"score":4.04},
         {"item":68,"score":3.81}
      ]
    }

The PreictedResult.java file has the class for the preceding conversion:

    package org.template.recommendation;

    import java.io.Serializable;
    import java.util.List;

    public class PredictedResult implements Serializable{
      private final List<ItemScore> itemScores;

      public PredictedResult(List<ItemScore> itemScores) {
        this.itemScores = itemScores;
      }

      public List<ItemScore> getItemScores() {
        return itemScores;
      }

      @Override
      public String toString() {
        return "PredictedResult{" +
        "itemScores=" + itemScores +
        '}';
      }
    } 

The Scala version of the code uses the case class, and you can define the query class and the predictedResult class as a case class:

    package MyRecommedationScala

    import org.apache.predictionio.controller.IEngineFactory
    import org.apache.predictionio.controller.Engine

    case class Query(
      user: String,
      num: Int,
      categories: Option[Set[String]],
      whiteList: Option[Set[String]],
      blackList: Option[Set[String]]
    ) extends Serializable

    case class PredictedResult(
    itemScores: Array[ItemScore]
    ) extends Serializable

    case class ItemScore(
    item: String,
    score: Double
    ) extends Serializable

    object ECommerceRecommendationEngine extends IEngineFactory {
      def apply() = {
        new Engine(
        classOf[DataSource],
        classOf[Preparator],
        Map("ecomm" -> classOf[ECommAlgorithm]),
        classOf[Serving])
      }
    }