Associate a Product Recommendation (PR) model with a query pipeline — Coveo Merchandising Hub

This is for:

Merchandising Hub Implementer System Administrator Developer

In this article

Step 1: Create the query pipeline conditions:
Step 2: Associate a PR model with the query pipeline for recommendations:

Instructions in this article are only relevant for Coveo for Commerce implementations that leverage the Coveo Merchandising Hub (CMH) to manage recommendation slot configurations. If you’re associating a PR model with a query pipeline that serves a recommendation interface that isn’t configured through the CMH, refer to the generic PR model association instructions.

When associating a PR model with the query pipeline for recommendations, you must configure an association for each of the CMH recommendation strategies you want to use in your storefront. For example, if you use the Most purchased and Viewed together CMH strategies in your storefront, you must configure an association for each of these strategies.

This process involves:

Creating query pipeline conditions to trigger the proper association based on the strategy used in the storefront.
Associating the model with the query pipeline for recommendations.

Step 1: Create the query pipeline conditions:

You must create a query pipeline condition for each of the CMH recommendation strategies you want to use in your storefront.

The following table provides the conditions you must create based on the CMH strategy you want to use in your storefront:

CMH strategy Query pipeline condition

CMH strategy	Query pipeline condition
Viewed together	`Recommendation is viewed_together`
Purchased together	`Recommendation is bought_together`
Intent-aware	`Recommendation is session_recommender_single_seed`
Most viewed	`Recommendation is popular_viewed`
Most purchased	`Recommendation is popular_bought`
Recently viewed	`Recommendation is recently_viewed`
Recently purchased	`Recommendation is recently_bought`
Purchased with recently purchased	`Recommendation is bought_with_recently_bought`
Buy again	`Recommendation is buy_again`

Viewed together

Recommendation is viewed_together

Purchased together

Recommendation is bought_together

Intent-aware

Recommendation is session_recommender_single_seed

Step 2: Associate a PR model with the query pipeline for recommendations:

To associate a PR model with a query pipeline, use the JSON editor of the Query Pipelines (platform-ca | platform-eu | platform-au) page in the Coveo Administration Console.

In the Coveo Administration Console, access the Query Pipelines page.
In the Query Pipelines page, click the query pipeline you configured for recommendations, and then click Edit components.
Click the Machine Learning tab, and then click the ellipsis in the upper right part of the page to select Associate a model in JSON view.

In the JSON editor, add the following configuration, replacing the placeholder values with your data, and then click Associate model.

{
  "modelId": "<MY_MODEL_ID>", 
  "condition": "<MY_CONDITION_ID>", 
  "maxRecommendations": 50,
  "cacheMaximumAge": "<CACHE_MAXIMUM_AGE>", 
  "exclusive": true,
  "customQueryParameters": {
    "strategy": "<STRATEGY_VALUE>" 
  },
  "useAdvancedConfiguration": true
}

Replace <MY_MODEL_ID> with the ID of the model you want to associate with the query pipeline. See Review model information to learn how to find the model ID.

Replace <MY_CONDITION_ID> with the ID of the query pipeline condition you created for the strategy you want to associate the model with.

When associating an SBPR model, we recommend adjusting the default value of PT10S to a shorter duration between 0 and 5 seconds (that is, PT0S to PT5S). This adjustment ensures the model responds more quickly to user behavior. Note that the value must be in the ISO-8601 format.

Replace <STRATEGY_VALUE> with the the strategy you want to associate the model with. The following table provides the strategy values for each CMH strategy:

CMH strategy Value to use

Viewed together

frequentViewed

Purchased together

frequentBought

Intent-aware

sessionRecommenderWithFallbackEnriched