# Source Ranking Issues ## The Problem Relevant documents ranked too low in retrieval results, causing less relevant or outdated sources to appear first and influence the AI response. ### Symptoms * ❌ Best answer in position 15, not top 3 * ❌ Generic content outranks specific * �tml Language Models over GPT-4) → Domain-adapted: Continues training on your docs→ Better understands your terminology → Returns your specific results first Cost: → Fine-tuning: $1000-5000 → Ongoing: Hosting custom model → Worth it for large-scale or specialized domains ``` --- ## How to Solve **Implement reranking (Cohere Rerank, cross-encoder) after initial retrieval + boost recent documents with recency scoring + assign source authority weights + use domain-adapted embeddings or fine-tuning + increase K at retrieval, then rerank top-20 to top-5 + test ranking quality with NDCG@K metric + adjust boost weights empirically (not just guessing).** See [Ranking Optimization](../accuracy/source-priority.md). ``` --- # Agent Instructions: Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://help.twig.so/rag-scenarios-and-solutions/accuracy/source-priority.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.