Pre-trained neural language models, such as BERT, fine-tuned for text ranking, have demonstrated remarkable effectiveness compared to baseline text ranking methods when evaluating the models in-domain. (View Highlight)
Information retrieval (IR) evaluation is the process of measuring the effectiveness of an information retrieval system (View Highlight)
standard IR metrics such as nDCG@10, Precision@10, and Recall@100. (View Highlight)
