RAG glossary

Prev Next

Available in VPC

To use RAG service, you must familiarize yourself with a few key terms. The glossary and its description are as follows:

Object Storage buckets

A space in storage that is allocated and holds a value or is referenced by an identifier (such as a variable, data structure, function, or method). In the Object Storage service, a bucket is the top-level directory that groups related objects together.

Index

The process of training data to build an AI-based search service in the RAG service

Chunking method

A method of splitting data into segments

Overlap

Refers to the overlap ratio between individual segments in the search results of chunked data

Clova Studio

A hyperscale AI development tool provided by NAVER Cloud Platform

API Key

An API key issued within the RAG service to validate service access; separate from the NAVER Cloud Platform-issued API key used for integrating other products

Prompt

An instruction-like prompt designed to elicit a response from an AI model

Parameter configuration

Token-related configuration values used in the RAG service

Note

For more information on terms, see glossary on the NAVER Cloud Platform portal.