Classic information retrieval collection with aeronautics documents, queries, and relevance judgments. Phase 1 indexes the 1,400-document collection.
Retail Search
ARCH-0.5 · cross-domain core
Mission-driven search experiments
Each phase gets its own focused workspace. The home page stays small: status, purpose, metrics, and links to the phase details.
Cranfield Foundation
Production-shaped OpenSearch BM25 baseline with public search, explain, and evaluation evidence.
Cross-Domain Validation (BEIR)
Test whether Phase 1 techniques transfer across domains. Result: BGE hybrid is universal (the ARCH-0.5 core); keyword rerankers are domain-conditional; learning-to-rank is portfolio.
Retail Relevance (Amazon ESCI)
Product search over 1.2M Amazon products with structured fields. The BGE hybrid reaches parity with the published baseline (0.8468 vs 0.8503) and understands intent keyword search cannot.
Behavioral Ranking
Introduce behavior signals only after earlier relevance gates justify the complexity.
Dataset references
Heterogeneous IR benchmark. Phase 2 baselined all 15 public datasets (BM25 avg nDCG@10 0.4261, on the published ~0.43) and classified techniques cross-domain: the BGE hybrid is universal (ARCH-0.5 core), keyword rerankers are domain-conditional, learning-to-rank is portfolio.
Product-search benchmark with Exact, Substitute, Complement, and Irrelevant relevance labels for query-product pairs.
Future phase. No public behavior dataset is selected yet; this phase will document the source, privacy boundary, and evaluation protocol before implementation.