P@k (Precision at k)
Manning, Raghavan, Schütze (2008) IIR §8.4
Open citationThis audit tests Search.Index.CamelCaseSplitter (#77), the cupertino-specific mechanism that expands CamelCase identifiers like LazyVGrid into recall-aiding components {Lazy, VGrid, Grid} stored in the docs_fts.symbol_components column. The column carries BM25F weight 1.5 (one-tenth of docs_fts.symbols at 5.0) so an exact identifier match wins over a fragment match, but a query for a fragment alone should still surface pages whose symbols contain it.
Each card opens its own page. The headline and charts above are all you need at a glance; the cards are for the why and how.
Every one of the 20 fragment queries returned a fragment-containing URI at rank 1. The symbol_components mechanism is doing its job at the top of the ranking.
Read details →The symbol_components column does what #77 designed it to do: a query for a CamelCase fragment surfaces pages whose symbols contain that fragment as a component.
Read details →20 fragment queries. For each: run cupertino search "<fragment>" --limit 10, extract top-10 URIs, apply a loose regex that matches URIs whose path contains the fragment as a substring in any segment.
Read details →Four of eight Phase 1.x classes from §1.4 now have documented baselines. Four remain: C (acronym), G (prose), H (symbol-attribute), and Phase 1.7 agent-end-to-end.
Read details →Every metric and method this audit relies on, with a link to the foundational source. Auto-collected from the audit text.
Manning, Raghavan, Schütze (2008) IIR §8.4
Open citationVoorhees (1999), TREC-8 QA Report
Open citation