TREC-DL

The TREC-DL dataset is a benchmark used to evaluate document ranking models, containing query-document pairs for assessing the effectiveness of ranking algorithms.

Papers using TREC-DL (3)

Supervised Contrastive Learning Approach for Contextual Ranking2022Data Augmentation for Sample Efficient and Robust Document Ranking2023Few-shot Prompting for Pairwise Ranking: An Effective Non-Parametric Retrieval Model2024