This is documentation for v2 of the Kotlin API clients, which is not the latest version. To see the documentation for the latest version, see Kotlin v3.

Required API Key: any key with the search ACL

Method signature
recommendClient.getRecommendations(requests: List<RecommendationsQuery>)

About this method #

Get recommendations from any Algolia recommendation model.

To get recommendations for specific models, use the specific methods instead—for example, getRelatedProducts, getTrendingItems, or getFrequentlyBoughtTogether.

Examples #

1
2
3
4
5
6
7
val request = RecommendationsQuery(
    indexName = IndexName("yourIndexName"),
    objectID = ObjectID("yourObjectID"),
    model = RecommendationModel.BoughtTogether,
)

val recommendations = recommendClient.getRecommendations(requests = listOf(request))

Parameters #

requests #
type: array of request object
Required

List of request objects.

requests âž” request object #

The parameters for the request body depend on the requested #{model}. The following parameters are common to all models.

indexName #
type: string
Required

Name of the index.

model #
type: "related-products" | "bought-together" | "trending-items" | "trending-facets"
Required

Name of the recommendation model to use.

threshold #
type: number
Required

Threshold for the recommendations confidence score (between 0 and 100). Only recommendations with a greater score are returned.

maxRecommendations #
type: integer

Number of recommendations to retrieve. Depending on the available recommendations and the other request parameters, the actual number of recommendations may be lower than that. If you don’t set maxRecommendations or set it to 0, all matching recommendations are returned, and no fallback request is performed.

requests âž” request object with bought-together #

{
  indexName: "YourIndexName",
  model: "bought-together",
  threshold: 0,
  maxRecommendations: 10,
  objectID: "an objectID",
  queryParameters: { query parameters },
}

Get recommendations from the ‘bought-together’ model given an object ID.

objectID #
type: string
Required

Object ID for which to get recommendations.

queryParameters #
type: object
Optional

Search parameters for filtering the recommendations.

requests âž” request object with related-products #

{
  indexName: "YourIndexName",
  model: "related-products",
  threshold: 0,
  maxRecommendations: 10,
  objectID: "an objectID",
  queryParameters: { query parameters },
  fallbackParameters: { query parameters }
}

Get recommendations from the ‘related-products’ model given an object ID.

objectID #
type: string
Required

Object ID for which to get recommendations.

queryParameters #
type: object
Optional

Search parameters for filtering the recommendations.

fallbackParameters #
type: object
Optional

Search parameters to use as fallback when there are no recommendations.

requests âž” request object with trending-items #

{
  indexName: "YourIndexName",
  model: "trending-items",
  threshold: 0,
  maxRecommendations: 10,
  facetName: "a facet attribute name",
  facetValue: "a facet value",
  queryParameters: { query parameters },
  fallbackParameters: { query parameters }
}

Fetch trending items, either globally or only those matching a specific facet. The following parameters are specific to model ‘trending-items’. If no facet is provided, the query fetches globally trending items.

facetName #
type: string
Optional

Facet attribute for which to get recommendations. This parameter must be used together with facetValue.

facetValue #
type: string
Optional

Facet value for which to get recommendations for. This parameter must be used along with facetName.

queryParameters #
type: object
Optional

Search parameters for filtering the recommendations.

fallbackParameters #
type: object
Optional

Search parameters to use as fallback when there are no recommendations.

requests âž” request object with trending-facets #

{
  indexName: "YourIndexName",
  model: "trending-facets",
  threshold: 0,
  maxRecommendations: 10,
  facetName: "a facet attribute name",
}

Fetch trending facet values given a specific facet attribute. The following parameters are specific to model ‘trending-facets’.

facetName #
type: string
Required

Facet attribute for which to get recommendations.

Response #

This section shows the JSON response returned by the API. Each API client encapsulates this response inside objects specific to the programming language, so that the actual response might be different. You can view the response by using the getLogs method. Don’t rely on the order of attributes in the response, as JSON doesn’t guarantee the ordering of keys in objects.

JSON format#

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
{
  "results": [
    {
      "hits": [
        {
          "_highlightResult": {
            "category": {
              "matchLevel": "none",
              "matchedWords": [],
              "value": "Men - T-Shirts"
            },
            "image_link": {
              "matchLevel": "none",
              "matchedWords": [],
              "value": "https://example.org/image/D05927-8161-111-F01.jpg"
            },
            "name": {
              "matchLevel": "none",
              "matchedWords": [],
              "value": "Jirgi Half-Zip T-Shirt"
            }
          },
          "_score": 32.72,
          "category": "Men - T-Shirts",
          "image_link": "https://example.org/image/D05927-8161-111-F01.jpg",
          "name": "Jirgi Half-Zip T-Shirt",
          "objectID": "D05927-8161-111",
          "position": 105,
          "url": "men/t-shirts/d05927-8161-111"
        }
      ],
      "processingTimeMS": 1,
    }
  ]
}
Field Description
results #
array of result

List of results in the order they were submitted, one per query.

{
  "results": [
    {
      "hits": [
        {
          ...,
          _score: 32.72
        }
      ],
    },
  ]
}

results âž” result #

The results object contains the same hits object as in the search method augmented by a _score attribute.

Field Description
_score #
number

Confidence score of the recommended item. The closer it is to 100, the more relevant.

Did you find this page helpful?
Kotlin API clients v2