This Recommendations AI documentation references the Recommendations console. We recommend switching to the Retail console and using the Retail documentation, which documents Recommendations AI, the Retail console, and Retail Search.

If you are using the v1beta version of Recommendations AI, migrate to the Retail API version.

REST Resource: projects.locations.catalogs.servingConfigs

Resource: ServingConfig

Configures metadata that is used to generate serving time results (e.g. search results or recommendation predictions). The ServingConfig is passed in the search and predict request and together with the Catalog.default_branch, generates results.

JSON representation 
{
  "name": string,
  "displayName": string,
  "modelId": string,
  "priceRerankingLevel": string,
  "facetControlIds": [
    string
  ],
  "dynamicFacetSpec": {
    object (DynamicFacetSpec)
  },
  "boostControlIds": [
    string
  ],
  "filterControlIds": [
    string
  ],
  "redirectControlIds": [
    string
  ],
  "twowaySynonymsControlIds": [
    string
  ],
  "onewaySynonymsControlIds": [
    string
  ],
  "doNotAssociateControlIds": [
    string
  ],
  "replacementControlIds": [
    string
  ],
  "ignoreControlIds": [
    string
  ],
  "diversityLevel": string,
  "enableCategoryFilterLevel": string,
  "solutionTypes": [
    enum (SolutionType)
  ]
}
Fields
name

string

Immutable. Fully qualified name projects/*/locations/global/catalogs/*/servingConfig/*
displayName

string

Required. The human readable serving config display name. Used in Retail UI.

This field must be a UTF-8 encoded string with a length limit of 128 characters. Otherwise, an INVALID_ARGUMENT error is returned.
modelId

string

The id of the model to use at serving time. Currently only RecommendationModels are supported: https://cloud.google.com/retail/recommendations-ai/docs/create-models Can be changed but only to a compatible model (e.g. others-you-may-like CTR to others-you-may-like CVR).

Required when solutionTypes is SOLUTION_TYPE_RECOMMENDATION.
priceRerankingLevel

string

How much price ranking we want in serving results. Price reranking causes product items with a similar recommendation probability to be ordered by price, with the highest-priced items first. This setting could result in a decrease in click-through and conversion rates. Allowed values are:

  • 'no-price-reranking'
  • 'low-price-raranking'
  • 'medium-price-reranking'
  • 'high-price-reranking'

If not specified, we choose default based on model type. Default value: 'no-price-reranking'.

Can only be set if solutionTypes is SOLUTION_TYPE_RECOMMENDATION.
facetControlIds[]

string

Facet specifications for faceted search. If empty, no facets are returned. The ids refer to the ids of Control resources with only the Facet control set. These controls are assumed to be in the same Catalog as the ServingConfig. A maximum of 100 values are allowed. Otherwise, an INVALID_ARGUMENT error is returned.

Can only be set if solutionTypes is SOLUTION_TYPE_SEARCH.
dynamicFacetSpec

object (DynamicFacetSpec)

The specification for dynamically generated facets. Notice that only textual facets can be dynamically generated.

Can only be set if solutionTypes is SOLUTION_TYPE_SEARCH.
boostControlIds[]

string

Condition boost specifications. If a product matches multiple conditions in the specifications, boost scores from these specifications are all applied and combined in a non-linear way. Maximum number of specifications is 100.

Notice that if both ServingConfig.boost_control_ids and SearchRequest.boost_spec are set, the boost conditions from both places are evaluated. If a search request matches multiple boost conditions, the final boost score is equal to the sum of the boost scores from all matched boost conditions.

Can only be set if solutionTypes is SOLUTION_TYPE_SEARCH.
filterControlIds[]

string

Condition filter specifications. If a product matches multiple conditions in the specifications, filters from these specifications are all applied and combined via the AND operator. Maximum number of specifications is 100.

Can only be set if solutionTypes is SOLUTION_TYPE_SEARCH.
redirectControlIds[]

string

Condition redirect specifications. Only the first triggered redirect action is applied, even if multiple apply. Maximum number of specifications is 1000.

Can only be set if solutionTypes is SOLUTION_TYPE_SEARCH.
twowaySynonymsControlIds[]

string

Condition synonyms specifications. If multiple syonyms conditions match, all matching synonyms control in the list will execute. Order of controls in the list will not matter. Maximum number of specifications is 100.

Can only be set if solutionTypes is SOLUTION_TYPE_SEARCH.
onewaySynonymsControlIds[]

string

Condition oneway synonyms specifications. If multiple oneway synonyms conditions match, all matching oneway synonyms controls in the list will execute. Order of controls in the list will not matter. Maximum number of specifications is 100.

Can only be set if solutionTypes is SOLUTION_TYPE_SEARCH.
doNotAssociateControlIds[]

string

Condition do not associate specifications. If multiple do not associate conditions match, all matching do not associate controls in the list will execute. - Order does not matter. - Maximum number of specifications is 100.

Can only be set if solutionTypes is SOLUTION_TYPE_SEARCH.
replacementControlIds[]

string

Condition replacement specifications. - Applied according to the order in the list. - A previously replaced term can not be re-replaced. - Maximum number of specifications is 100.

Can only be set if solutionTypes is SOLUTION_TYPE_SEARCH.
ignoreControlIds[]

string

Condition ignore specifications. If multiple ignore conditions match, all matching ignore controls in the list will execute. - Order does not matter. - Maximum number of specifications is 100.

Can only be set if solutionTypes is SOLUTION_TYPE_SEARCH.
diversityLevel

string

How much diversity to use in recommendation model results e.g. 'medium-diversity' or 'high-diversity'. Currently supported values:

  • 'no-diversity'
  • 'low-diversity'
  • 'medium-diversity'
  • 'high-diversity'
  • 'auto-diversity'

If not specified, we choose default based on recommendation model type. Default value: 'no-diversity'.

Can only be set if solutionTypes is SOLUTION_TYPE_RECOMMENDATION.
enableCategoryFilterLevel

string

Whether to add additional category filters on the 'similar-items' model. If not specified, we enable it by default. Allowed values are:

  • 'no-category-match': No additional filtering of original results from the model and the customer's filters.
  • 'relaxed-category-match': Only keep results with categories that match at least one item categories in the PredictRequests's context item.
  • If customer also sends filters in the PredictRequest, then the results will satisfy both conditions (user given and category match).

Can only be set if solutionTypes is SOLUTION_TYPE_RECOMMENDATION.
solutionTypes[]

enum (SolutionType)

Required. Immutable. Specifies the solution types that a serving config can be associated with. Currently we support setting only one type of solution.

DynamicFacetSpec

The specifications of dynamically generated facets.

JSON representation 
{
  "mode": enum (Mode)
}
Fields
mode

enum (Mode)

Mode of the DynamicFacet feature. Defaults to Mode.DISABLED if it's unset.

Mode

Enum to control DynamicFacet mode

Enums
MODE_UNSPECIFIED Default value.
DISABLED Disable Dynamic Facet.
ENABLED Automatic mode built by Google Retail placements.search.

Methods

addControl

 Enables a Control on the specified ServingConfig.

create

 Creates a ServingConfig.

delete

 Deletes a ServingConfig.

get

 Gets a ServingConfig.

list

 Lists all ServingConfigs linked to this catalog.

patch

 Updates a ServingConfig.

predict

 Makes a recommendation prediction.

removeControl

 Disables a Control on the specified ServingConfig.
Performs a search.