theagilemonkeys/aifindr-search-eval

--- license: apache-2.0 task_categories: - text-retrieval language: - en - es tags: - e-commerce - fashion - search - evaluation size_categories: - 1K<n<10K --- # AIFindr Search — Evaluation Data Evaluation datasets for the AIFindr Search research project. Contains labeled query-product relevance judgments, evaluation queries, catalog vocabulary, and experiment results. ## Contents - **v5/**, **v6/**: Labeled datasets with ESCI-graded relevance judgments - **queries/**: Evaluation query sets (v4 base + v6 expanded) - **catalog/**: Product catalog vocabulary for filter generation - **filters/**: GPT-5.2 generated query filters and expansions - **results/**: Evaluation results and significance tests ## Usage ```python from datasets import load_dataset ds = load_dataset("theagilemonkeys/aifindr-search-eval", data_files="v6/labeled-dataset-v6.json") ``` Or with the companion download script: ```bash cd research && uv run python experiments/tools/download_data.py --repo eval ```