Custom Content Recommendation API Reference

All Sherpa.ai Custom Content Recommendation API endpoints will be served on https://recom.sherpa.ai/v2/recomm

This REST API is used to get user and item recommendations, based on the interactions between them, and lets you define a database composed of tables and add items to those tables. User identities can also be defined, in order to log interactions between users and items. This API also allows for the definition of both items and users, in order to provide sorting and filtering capabilities for queries. The recommender analyzes all this data and provides the following recommendations:

• Items to Users: the most relevant items for a user
• Users to Items: the most suitable users for an item
• Items to Items: the most similar items to a given item
• Users to Users: the most similar users to a given user

Getting Started

With the Sherpa.ai Custom Content Recommendation API, quickly and easily add all the data necessary to get the recommendations you need. Simply get an API key to authenticate all your requests. This key will allow you to manage your data within the platform and access all the recommendation endpoints that the recommender provides.

Preparation

• Add tables and their attributes. Then, start adding items to them.
• Define user attributes. Then, start adding users.
• Define interactions between users and items for each table.
• Add user-item interactions.

Consumption

Once all the basic data is defined, query for recommendations:

• Get items for a user to get the most relevant items for a user
• Get users for an item to get the most suitable users for an item
• Get items for an item to get the most similar items to a given item
• Get users for a user to get the most similar users to a given user