RecommendItemsToItem

Extends \Recombee\RecommApi\Requests\Request

Recommends set of items that are somehow related to one given item, *X*. Typical scenario is when user *A* is viewing *X*. Then you may display items to the user that he might be also interested in. Recommend items to item request gives you Top-N such items, optionally taking the target user *A* into account.

It is also possible to use POST HTTP method (for example in case of very long ReQL filter) - query parameters then become body parameters.

package

Default

Methods

Construct the request

__construct(string $item_id, string $target_user_id, integer $count, array $optional = array()) 
Throws
\Recombee\RecommApi\Requests\Exceptions\UnknownOptionalParameterException

UnknownOptionalParameterException if an unknown optional parameter is given in $optional

Arguments

$item_id

string

ID of the item for which the recommendations are to be generated.

$target_user_id

string

ID of the user who will see the recommendations. Specifying the targetUserId is beneficial because:

  • It makes the recommendations personalized
  • Allows the calculation of Actions and Conversions in the graphical user interface, as Recombee can pair the user who got recommendations and who afterwards viewed/purchased an item. If you insist on not specifying the user, pass null (None, nil, NULL etc. depending on language) to targetUserId. Do not create some special dummy user for getting recommendations, as it could mislead the recommendation models, and result in wrong recommendations. For anonymous/unregistered users it is possible to use for example their session ID.

$count

integer

Number of items to be recommended (N for the top-N recommendation).

$optional

array

Optional parameters given as an array containing pairs name of the parameter => value

  • Allowed parameters:
    • userImpact
      • Type: float
      • Description: If targetUserId parameter is present, the recommendations are biased towards the user given. Using userImpact, you may control this bias. For an extreme case of userImpact=0.0, the interactions made by the user are not taken into account at all (with the exception of history-based blacklisting), for userImpact=1.0, you'll get user-based recommendation. The default value is 0.
    • filter
      • Type: string
      • Description: Boolean-returning ReQL expression which allows you to filter recommended items based on the values of their attributes.
    • booster
      • Type: string
      • Description: Number-returning ReQL expression which allows you to boost recommendation rate of some items based on the values of their attributes.
    • cascadeCreate
      • Type: bool
      • Description: If item of given itemId or user of given targetUserId doesn't exist in the database, it creates the missing entity/entities and returns some (non-personalized) recommendations. This allows for example rotations in the following recommendations for the user of given targetUserId, as the user will be already known to the system.
    • scenario
      • Type: string
      • Description: Scenario defines a particular application of recommendations. It can be for example "homepage", "cart" or "emailing". You can see each scenario in the UI separately, so you can check how well each application performs. The AI which optimizes models in order to get the best results may optimize different scenarios separately, or even use different models in each of the scenarios.
    • returnProperties
      • Type: bool
      • Description: With returnProperties=true, property values of the recommended items are returned along with their IDs in a JSON dictionary. The acquired property values can be used for easy displaying of the recommended items to the user. Example response:
        {
        "recommId": "0c6189e7-dc1a-429a-b613-192696309361",
        "recomms":
        [
        {
        "id": "tv-178",
        "values": {
        "description": "4K TV with 3D feature",
        "categories":   ["Electronics", "Televisions"],
        "price": 342,
        "url": "myshop.com/tv-178"
        }
        },
        {
        "id": "mixer-42",
        "values": {
        "description": "Stainless Steel Mixer",
        "categories":   ["Home & Kitchen"],
        "price": 39,
        "url": "myshop.com/mixer-42"
        }
        }
        ]
        }
    • includedProperties
      • Type: array
      • Description: Allows to specify, which properties should be returned when returnProperties=true is set. The properties are given as a comma-separated list. Example response for includedProperties=description,price:
        {
        "recommId": "6842c725-a79f-4537-a02c-f34d668a3f80",
        "recomms":
        [
        {
        "id": "tv-178",
        "values": {
        "description": "4K TV with 3D feature",
        "price": 342
        }
        },
        {
        "id": "mixer-42",
        "values": {
        "description": "Stainless Steel Mixer",
        "price": 39
        }
        }
        ]
        }
    • diversity
      • Type: float
      • Description: Expert option Real number from [0.0, 1.0] which determines how much mutually dissimilar should the recommended items be. The default value is 0.0, i.e., no diversification. Value 1.0 means maximal diversification.
    • minRelevance
      • Type: string
      • Description: Expert option If the targetUserId is provided: Specifies the threshold of how much relevant must the recommended items be to the user. Possible values one of: "low", "medium", "high". The default value is "low", meaning that the system attempts to recommend number of items equal to count at any cost. If there are not enough data (such as interactions or item properties), this may even lead to bestseller-based recommendations to be appended to reach the full count. This behavior may be suppressed by using "medium" or "high" values. In such case, the system only recommends items of at least the requested relevancy, and may return less than count items when there is not enough data to fulfill it.
    • rotationRate
      • Type: float
      • Description: Expert option If the targetUserId is provided: If your users browse the system in real-time, it may easily happen that you wish to offer them recommendations multiple times. Here comes the question: how much should the recommendations change? Should they remain the same, or should they rotate? Recombee API allows you to control this per-request in backward fashion. You may penalize an item for being recommended in the near past. For the specific user, rotationRate=1 means maximal rotation, rotationRate=0 means absolutely no rotation. You may also use, for example rotationRate=0.2 for only slight rotation of recommended items.
    • rotationTime
      • Type: float
      • Description: Expert option If the targetUserId is provided: Taking rotationRate into account, specifies how long time it takes to an item to recover from the penalization. For example, rotationTime=7200.0 means that items recommended less than 2 hours ago are penalized.
    • expertSettings
      • Type:
      • Description: Dictionary of custom options.

Get body parameters

getBodyParameters() : array
inherited abstract

Response

array

Values of body parameters (name of parameter => value of the parameter)

Returns true if HTTPS must be chosen over HTTP for this request

getEnsureHttps() : boolean
inherited

Response

boolean

true if HTTPS must be chosen

Get used HTTP method

getMethod() : static
inherited abstract

Response

static

Used HTTP method

Get URI to the endpoint

getPath() : string
inherited abstract

Response

string

URI to the endpoint

Get query parameters

getQueryParameters() : array
inherited abstract

Response

array

Values of query parameters (name of parameter => value of the parameter)

Get request timeout

getTimeout() : integer
inherited

Response

integer

Request timeout in milliseconds

Sets if HTTPS must be chosen over HTTP for this request

setEnsureHttps( $ensure_https) 
inherited

Arguments

$ensure_https

Sets request timeout

setTimeout( $timeout) 
inherited

Arguments

$timeout

Properties

ID of the item for which the recommendations are to be generated.

item_id : string
var

ID of the item for which the recommendations are to be generated.

Type(s)

string

ID of the user who will see the recommendations. Specifying the *targetUserId* is beneficial because: * It makes the recommendations personalized * Allows the calculation of Actions and Conversions in the graphical user interface, as Recombee can pair the user who got recommendations and who afterwards viewed/purchased an item. If you insist on not specifying the user, pass `null` (`None`, `nil`, `NULL` etc. depending on language) to *targetUserId*. Do not create some special dummy user for getting recommendations, as it could mislead the recommendation models, and result in wrong recommendations. For anonymous/unregistered users it is possible to use for example their session ID.

target_user_id : string
var

ID of the user who will see the recommendations. Specifying the targetUserId is beneficial because:

  • It makes the recommendations personalized
  • Allows the calculation of Actions and Conversions in the graphical user interface, as Recombee can pair the user who got recommendations and who afterwards viewed/purchased an item. If you insist on not specifying the user, pass null (None, nil, NULL etc. depending on language) to targetUserId. Do not create some special dummy user for getting recommendations, as it could mislead the recommendation models, and result in wrong recommendations. For anonymous/unregistered users it is possible to use for example their session ID.

Type(s)

string

Number of items to be recommended (N for the top-N recommendation).

count : integer
var

Number of items to be recommended (N for the top-N recommendation).

Type(s)

integer

If *targetUserId* parameter is present, the recommendations are biased towards the user given. Using *userImpact*, you may control this bias. For an extreme case of `userImpact=0.0`, the interactions made by the user are not taken into account at all (with the exception of history-based blacklisting), for `userImpact=1.0`, you'll get user-based recommendation. The default value is `0`.

user_impact : float
var

If targetUserId parameter is present, the recommendations are biased towards the user given. Using userImpact, you may control this bias. For an extreme case of userImpact=0.0, the interactions made by the user are not taken into account at all (with the exception of history-based blacklisting), for userImpact=1.0, you'll get user-based recommendation. The default value is 0.

Type(s)

float

Boolean-returning [ReQL](https://docs.recombee.com/reql.html) expression which allows you to filter recommended items based on the values of their attributes.

filter : string
var

Boolean-returning ReQL expression which allows you to filter recommended items based on the values of their attributes.

Type(s)

string

Number-returning [ReQL](https://docs.recombee.com/reql.html) expression which allows you to boost recommendation rate of some items based on the values of their attributes.

booster : string
var

Number-returning ReQL expression which allows you to boost recommendation rate of some items based on the values of their attributes.

Type(s)

string

If item of given *itemId* or user of given *targetUserId* doesn't exist in the database, it creates the missing entity/entities and returns some (non-personalized) recommendations. This allows for example rotations in the following recommendations for the user of given *targetUserId*, as the user will be already known to the system.

cascade_create : boolean
var

If item of given itemId or user of given targetUserId doesn't exist in the database, it creates the missing entity/entities and returns some (non-personalized) recommendations. This allows for example rotations in the following recommendations for the user of given targetUserId, as the user will be already known to the system.

Type(s)

boolean

Scenario defines a particular application of recommendations. It can be for example "homepage", "cart" or "emailing". You can see each scenario in the UI separately, so you can check how well each application performs. The AI which optimizes models in order to get the best results may optimize different scenarios separately, or even use different models in each of the scenarios.

scenario : string
var

Scenario defines a particular application of recommendations. It can be for example "homepage", "cart" or "emailing". You can see each scenario in the UI separately, so you can check how well each application performs. The AI which optimizes models in order to get the best results may optimize different scenarios separately, or even use different models in each of the scenarios.

Type(s)

string

With `returnProperties=true`, property values of the recommended items are returned along with their IDs in a JSON dictionary. The acquired property values can be used for easy displaying of the recommended items to the user. Example response: ``` { "recommId": "0c6189e7-dc1a-429a-b613-192696309361", "recomms": [ { "id": "tv-178", "values": { "description": "4K TV with 3D feature", "categories": ["Electronics", "Televisions"], "price": 342, "url": "myshop.com/tv-178" } }, { "id": "mixer-42", "values": { "description": "Stainless Steel Mixer", "categories": ["Home & Kitchen"], "price": 39, "url": "myshop.com/mixer-42" } } ] } ```

return_properties : boolean
var

With returnProperties=true, property values of the recommended items are returned along with their IDs in a JSON dictionary. The acquired property values can be used for easy displaying of the recommended items to the user. Example response:

  {
    "recommId": "0c6189e7-dc1a-429a-b613-192696309361",
    "recomms":
      [
        {
          "id": "tv-178",
          "values": {
            "description": "4K TV with 3D feature",
            "categories":   ["Electronics", "Televisions"],
            "price": 342,
            "url": "myshop.com/tv-178"
          }
        },
        {
          "id": "mixer-42",
          "values": {
            "description": "Stainless Steel Mixer",
            "categories":   ["Home & Kitchen"],
            "price": 39,
            "url": "myshop.com/mixer-42"
          }
        }
      ]
  }

Type(s)

boolean

Allows to specify, which properties should be returned when `returnProperties=true` is set. The properties are given as a comma-separated list. Example response for `includedProperties=description,price`: ``` { "recommId": "6842c725-a79f-4537-a02c-f34d668a3f80", "recomms": [ { "id": "tv-178", "values": { "description": "4K TV with 3D feature", "price": 342 } }, { "id": "mixer-42", "values": { "description": "Stainless Steel Mixer", "price": 39 } } ] } ```

included_properties : array
var

Allows to specify, which properties should be returned when returnProperties=true is set. The properties are given as a comma-separated list. Example response for includedProperties=description,price:

  {
    "recommId": "6842c725-a79f-4537-a02c-f34d668a3f80",
    "recomms": 
      [
        {
          "id": "tv-178",
          "values": {
            "description": "4K TV with 3D feature",
            "price": 342
          }
        },
        {
          "id": "mixer-42",
          "values": {
            "description": "Stainless Steel Mixer",
            "price": 39
          }
        }
      ]
  }

Type(s)

array

**Expert option** Real number from [0.0, 1.0] which determines how much mutually dissimilar should the recommended items be. The default value is 0.0, i.e., no diversification. Value 1.0 means maximal diversification.

diversity : float
var

Expert option Real number from [0.0, 1.0] which determines how much mutually dissimilar should the recommended items be. The default value is 0.0, i.e., no diversification. Value 1.0 means maximal diversification.

Type(s)

float

**Expert option** If the *targetUserId* is provided: Specifies the threshold of how much relevant must the recommended items be to the user. Possible values one of: "low", "medium", "high". The default value is "low", meaning that the system attempts to recommend number of items equal to *count* at any cost. If there are not enough data (such as interactions or item properties), this may even lead to bestseller-based recommendations to be appended to reach the full *count*. This behavior may be suppressed by using "medium" or "high" values. In such case, the system only recommends items of at least the requested relevancy, and may return less than *count* items when there is not enough data to fulfill it.

min_relevance : string
var

Expert option If the targetUserId is provided: Specifies the threshold of how much relevant must the recommended items be to the user. Possible values one of: "low", "medium", "high". The default value is "low", meaning that the system attempts to recommend number of items equal to count at any cost. If there are not enough data (such as interactions or item properties), this may even lead to bestseller-based recommendations to be appended to reach the full count. This behavior may be suppressed by using "medium" or "high" values. In such case, the system only recommends items of at least the requested relevancy, and may return less than count items when there is not enough data to fulfill it.

Type(s)

string

**Expert option** If the *targetUserId* is provided: If your users browse the system in real-time, it may easily happen that you wish to offer them recommendations multiple times. Here comes the question: how much should the recommendations change? Should they remain the same, or should they rotate? Recombee API allows you to control this per-request in backward fashion. You may penalize an item for being recommended in the near past. For the specific user, `rotationRate=1` means maximal rotation, `rotationRate=0` means absolutely no rotation. You may also use, for example `rotationRate=0.2` for only slight rotation of recommended items.

rotation_rate : float
var

Expert option If the targetUserId is provided: If your users browse the system in real-time, it may easily happen that you wish to offer them recommendations multiple times. Here comes the question: how much should the recommendations change? Should they remain the same, or should they rotate? Recombee API allows you to control this per-request in backward fashion. You may penalize an item for being recommended in the near past. For the specific user, rotationRate=1 means maximal rotation, rotationRate=0 means absolutely no rotation. You may also use, for example rotationRate=0.2 for only slight rotation of recommended items.

Type(s)

float

**Expert option** If the *targetUserId* is provided: Taking *rotationRate* into account, specifies how long time it takes to an item to recover from the penalization. For example, `rotationTime=7200.0` means that items recommended less than 2 hours ago are penalized.

rotation_time : float
var

Expert option If the targetUserId is provided: Taking rotationRate into account, specifies how long time it takes to an item to recover from the penalization. For example, rotationTime=7200.0 means that items recommended less than 2 hours ago are penalized.

Type(s)

float

Dictionary of custom options.

expert_settings : 
var

Dictionary of custom options.

Type(s)

Array containing values of optional parameters

optional : array
var

Array containing values of optional parameters

Type(s)

array

Timeout of the request in milliseconds

timeout : integer
inherited
var

Timeout of the request in milliseconds

Type(s)

integer

Sets if the HTTPS must be chosen over HTTP for this request

ensure_https : boolean
inherited
var

Sets if the HTTPS must be chosen over HTTP for this request

Type(s)

boolean