DataGEMS Query Recommender Api

This is the documentation site for the DataGEMS Query Recommender Api. The service is part of the wider DataGEMS platform. This is part of the overall platform documentation.

The Query Recommender Api suggests what a user might want to ask next. Given the natural language query the user has just submitted, it returns up to 3 short follow-up queries, personalised with the user's interest keywords as maintained by the User Profiler service. The suggestions are generated by a Large Language Model (LLM) served by the DataGEMS LLM endpoint.

A typical exchange looks like this:

POST /api/v1/recommend-next-queries
Authorization: Bearer <token>
Content-Type: application/json

{
  "current_query": "What is there to do in Verona tonight?",
  "context": { "user_id": "user1" }
}

{
  "next_queries": [
    "Concerts in Verona this weekend",
    "Opera at the Arena di Verona",
    "Late night restaurants in Verona"
  ]
}

You can use the menu options on the left to navigate through the available documentation. You may be interested to jump directly to an Architecture Overview, see the available Onboarding Material, read about some Api Constructs, investigate how to interpret Status & Error Codes or jump straight in to the OpenAPI Reference. You can find the repository for the service here: code repository.

For any questions, you can consult the FAQ, check if there is a relevant issue answering your question or contact the DataGEMS Help Desk.