<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Information Retrieval on tomato blog</title><link>https://yuhi-sa.github.io/en/tags/information-retrieval/</link><description>Recent content in Information Retrieval on tomato blog</description><generator>Hugo</generator><language>en</language><lastBuildDate>Sun, 22 Nov 2026 23:07:39 +0900</lastBuildDate><atom:link href="https://yuhi-sa.github.io/en/tags/information-retrieval/index.xml" rel="self" type="application/rss+xml"/><item><title>Evaluating RAG Retrieval in Python: Recall@k, MRR, and nDCG</title><link>https://yuhi-sa.github.io/en/posts/20261122_rag_retrieval_evaluation/1/</link><pubDate>Sun, 22 Nov 2026 00:00:00 +0900</pubDate><guid>https://yuhi-sa.github.io/en/posts/20261122_rag_retrieval_evaluation/1/</guid><description>&lt;p&gt;A reranker can find room for a direct answer at rank one even when the number of relevant results stays the same. Recall@k tracks missing evidence, MRR tracks the first relevant rank, and nDCG rewards placing the most relevant evidence near the top. We&amp;rsquo;ll use three queries and a small Python example to see where these metrics agree and where they diverge.&lt;/p&gt;</description></item><item><title>How to Evaluate RAG Chunking: Compare Chunk Size and Overlap in Python</title><link>https://yuhi-sa.github.io/en/posts/20261122_rag_chunking_evaluation/1/</link><pubDate>Sun, 22 Nov 2026 00:00:00 +0900</pubDate><guid>https://yuhi-sa.github.io/en/posts/20261122_rag_chunking_evaluation/1/</guid><description>&lt;p&gt;Give two RAG systems their top five chunks and the system with larger chunks can send more text to the LLM. To compare chunk size and overlap, label the evidence in the original documents and inspect what reaches the model under the same input token budget. A small Python example shows how to measure both missing evidence and duplicated text.&lt;/p&gt;</description></item></channel></rss>