A lighthouse-archive scrape keeps navigation chrome, click-here lines, and empty pages. What preprocessing step comes before any length plot or index build?
Select an answer to reveal the explanation.
Short Explanation
Navigation chrome, click-here lines, and empty pages leave the scrape before any length plot or index build. Strip boilerplate and empty documents as core text cleaning. A later safety filter, a RAG definition, and a NIM hop do not clean the corpus.
Full Explanation
Boilerplate and empty-document removal is core text cleaning. Navigation chrome and click-here lines should leave the corpus before a length plot or a vector index. This is not a Guardrails answer-time filter and not a RAG definition. Clean the scrape first.