Skip to content

REST Interface

Lucio Paiva edited this page Jun 1, 2015 · 2 revisions

The list below shows the main endpoints which are currently available, and their GET/POST parameters:

  • POST

    • process_activity_fastlane

      • env: the "environment", i.e., the alias for the client in barbante/config/mongodb.yml file;
      • user_id: the unique identifier of the user;
      • product_id: the unique identifier of the product;
      • activity_type: the type of the activity, which should be one of the types configured in barbante/config/barbante_<client_name>.yml file;
      • activity_date: the date/time of the activity, in ISO format
    • process_activity_slowlane

      • env: the "environment", i.e., the alias for the client in barbante/config/mongodb.yml file;
      • user_id: the unique identifier of the user;
      • product_id: the unique identifier of the product;
      • activity_type: the type of the activity, which should be one of the types configured in barbante/config/barbante_<client_name>.yml file;
      • activity_date: the date/time of the activity, in ISO format
    • process_product

      • env: the "environment", i.e., the alias for the client in barbante/config/mongodb.yml file;
      • product: a JSON document with all relevant product attributes, as configured in barbante/config/barbante_<client_name>.yml file;
  • process_impression

    • env: the "environment", i.e., the alias for the client in barbante/config/mongodb.yml file;
    • user_id: the unique identifier of the user;
    • product_id: the unique identifier of the product;
    • impression_date: the date/time of the activity, in ISO format
- delete_product
  - env: the "environment", i.e., the alias for the client in barbante/config/mongodb.yml file;
  - product_id: the unique identifier of the product
  • GET
    • recommend/<env>/<user_id>/<count_recommendations>/<algorithm>/<context_filter>
      • env: the "environment", i.e., the alias for the client in barbante/config/mongodb.yml file;
      • user_id: the unique identifier of the intended target user;
      • count_recommendations: the desired number of recommendations;
      • algorithm: the identification of the algorithm (UBCF, PBCF, CB, POP, HRChunks, HRRandom, HRVoting);
      • context_filter: an optional JSON document with product attributes to be matched (the supported product attributes are those marked "context_filter: true" in the PRODUCT_MODELS section in barbante/config/barbante_<client_name>.yml file)
Clone this wiki locally