Work

Improving Keyword Quality in a Production RAG Workflow

Year: 2026

Type: article

Abstract: This engineering white paper presents a bounded, product-local evaluation of keyword extraction and filtering in a production RAG workflow. It compares a reference YAKE configuration, an improved deterministic YAKE-v2 pipeline, a full LLM filter, and a narrowly scoped LLM fragment filter. The results support YAKE-v2 as the production default while retaining the narrow LLM policy as an off-by-default, fail-open option.

Source: Zenodo (CERN European Organization for Nuclear Research)

DOI: 10.5281/zenodo.22033277

Author: Mugade, Pratikraj

Language: English

Topic: Advanced Text Analysis Techniques

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: green

OpenAlex ID: W7203886700

FWCI:

Cites:

Cited by:

Related to: