Before You Begin

Prior to getting access to Rabbit Recommendations, each developer account must be approved. To get access developers need:

  • A paid subscription to one of IVA's core solutions.
  • Must work with IVA Sales to gain access.

Please note, Rabbit Recommendations data is not available on Free Trial subscriptions. For more information, contact Sales.

General Information

The Rabbit Recommendations content is an enrichment of the metadata in the Entertainment API with recommendations of similar Entertainment Programs.

Review the definitions and examples below to see how to work with the Rabbit Recommendations data. If you have any questions please reach out to our support team.


Related API Guides: Entertainment API

Open API/Swagger documentation can be found on the Rabbit Recommendations API page.


Rabbit Recommendations API Parameters

The Recommendations endpoint has a few parameters offering options to explore the Rabbit Recommendations results. Those parameters are detailed below:

MovieId Numerical Entertainment Id for the source Movie for which recommendations will be generated. Only MovieId or ShowId should be used but not both in one request.
ShowId Numerical Entertainment Id for the source Show for which recommendations will be generated. Only ShowId or MovieId should be used but not both in one request.
Type Type of programs to be included in recommendation calculation and results. "similar" will include Movies and Shows. "similiarShowsOnly" will return only Shows and "similarMoviesOnly" will return only Movies.

Rabbit Recommendations Response

An example of a Rabbit Recommendations response is below:

        {
            "recommendations": [
              {
                "explanations": [
                  {
                    "originator": "Similarity Engine",
                    "type": "feature",
                    "message": "Related through: genre Sci-Fi, genre Action-Adventure, Robert Downey Jr., Chris Evans, Scarlett Johansson, keyword superhero, keyword based on comic"
                  }
                ],
                "score": 0.6798871,
                "program": {
                  "Id": 126359,
                  "Type": "Movie"
                }
              },
              {
                "explanations": [
                  {
                    "originator": "Similarity Engine",
                    "type": "feature",
                    "message": "Related through: genre Sci-Fi, genre Action-Adventure, Chris Evans, Scarlett Johansson, Sebastian Stan, keyword superhero, keyword based on comic"
                  }
                ],
                "score": 0.66413033,
                "program": {
                  "Id": 8704,
                  "Type": "Movie"
                }
              },
        ...
      

There are several metadata fields that make up a Rabbit Recommendations object. They are described below:

program Object containing both an "Id" and "Type" field that represents the recommended Entertainment program. The "Id" and "Type" can be combined to form the full Entertainment ID (i.e. "Movie/126359").
score Decimal value between 0 and 1 representing quality of recommendation relative to other recommendations in response. (i.e. A Score of 1 would represent identical programs using the similarity engine.)
explanations Array of explanation objects to explain the origin of the recommendation.
originator Field in an explanation object describing the engine that was used to determine the recommendation. All recommendations use the "Similarity Engine" at this time.
type Field in an explanation object describing type of explanation. "reference" indicates a reference to the requested Entertainment Program, "feature" indicates detailed information about the requested Entertainment Program (enumerated in the "message" field), and "engine" indicates the most generic, highest-level recommendation.
message Field in an explanation object detailing the explanation. In "feature" explanation types, this will include specific metadata fields from the Entertainment Program (i.e. Genres, Contributor names, Keywords, etc.) that were used in making the recommendation.