Approximate Nearest Neighbor Negative Contrastive Learning for Dense Text Retrieval

Conducting text retrieval in a dense learned representation space has many intriguing advantages over sparse retrieval. Yet the effectiveness of dense retrieval (DR) often requires combination with sparse retrieval. In this paper, we identify that the main bottleneck is in the training mechanisms, where the negative instances used in training are not representative of the irrelevant documents in testing. This paper presents Approximate nearest neighbor Negative Contrastive Estimation (ANCE), a training mechanism that constructs negatives from an Approximate Nearest Neighbor (ANN) index of the corpus, which is parallelly updated with the learning process to select more realistic negative training instances. This fundamentally resolves the discrepancy between the data distribution used in the training and testing of DR. In our experiments, ANCE boosts the BERT-Siamese DR model to outperform all competitive dense and sparse retrieval baselines. It nearly matches the accuracy of sparse-retrieval-and-BERT-reranking using dot-product in the ANCE-learned representation space and provides almost 100x speed-up.

PDF Abstract ICLR 2021 PDF ICLR 2021 Abstract
Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Passage Retrieval Natural Questions ANCE Precision@20 81.9 # 6
Precision@100 87.5 # 7

Results from Other Papers

Task Dataset Model Metric Name Metric Value Rank Source Paper Compare
Zero-shot Text Search BEIR ANCE (Xiong et al., 2020) Avg. Accuracy 43.7 # 17


No methods listed for this paper. Add relevant methods here