Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

LADER: Log-Augmented DEnse Retrieval for Biomedical Literature Search

Qiao Jin, Ashley Shin, Zhiyong Lu

TL;DR

LADER introduces a plug-in mechanism that augments a dense biomedical retriever with click logs from similar training queries to improve literature search performance. By combining dense retrieval scores with log-augmented scores from related queries, LADER achieves state-of-the-art results on the TripClick benchmark, with substantial gains on HEAD queries and meaningful improvements across all query frequencies without additional training. The approach relies on a PubMedBERT-based dense backbone and a log-augmentation term controlled by a hyperparameter, demonstrating that query-based click signals can meaningfully enhance retrieval in short-query biomedical settings. While effective, LADER increases search latency due to cross-query retrieval and was evaluated on a single search engine, suggesting avenues for further efficiency and generalization exploration.

Abstract

Queries with similar information needs tend to have similar document clicks, especially in biomedical literature search engines where queries are generally short and top documents account for most of the total clicks. Motivated by this, we present a novel architecture for biomedical literature search, namely Log-Augmented DEnse Retrieval (LADER), which is a simple plug-in module that augments a dense retriever with the click logs retrieved from similar training queries. Specifically, LADER finds both similar documents and queries to the given query by a dense retriever. Then, LADER scores relevant (clicked) documents of similar queries weighted by their similarity to the input query. The final document scores by LADER are the average of (1) the document similarity scores from the dense retriever and (2) the aggregated document scores from the click logs of similar queries. Despite its simplicity, LADER achieves new state-of-the-art (SOTA) performance on TripClick, a recently released benchmark for biomedical literature retrieval. On the frequent (HEAD) queries, LADER largely outperforms the best retrieval model by 39% relative NDCG@10 (0.338 v.s. 0.243). LADER also achieves better performance on the less frequent (TORSO) queries with 11% relative NDCG@10 improvement over the previous SOTA (0.303 v.s. 0.272). On the rare (TAIL) queries where similar queries are scarce, LADER still compares favorably to the previous SOTA method (NDCG@10: 0.310 v.s. 0.295). On all queries, LADER can improve the performance of a dense retriever by 24%-37% relative NDCG@10 while not requiring additional training, and further performance improvement is expected from more logs. Our regression analysis has shown that queries that are more frequent, have higher entropy of query similarity and lower entropy of document similarity, tend to benefit more from log augmentation.

LADER: Log-Augmented DEnse Retrieval for Biomedical Literature Search

TL;DR

LADER introduces a plug-in mechanism that augments a dense biomedical retriever with click logs from similar training queries to improve literature search performance. By combining dense retrieval scores with log-augmented scores from related queries, LADER achieves state-of-the-art results on the TripClick benchmark, with substantial gains on HEAD queries and meaningful improvements across all query frequencies without additional training. The approach relies on a PubMedBERT-based dense backbone and a log-augmentation term controlled by a hyperparameter, demonstrating that query-based click signals can meaningfully enhance retrieval in short-query biomedical settings. While effective, LADER increases search latency due to cross-query retrieval and was evaluated on a single search engine, suggesting avenues for further efficiency and generalization exploration.

Abstract

Queries with similar information needs tend to have similar document clicks, especially in biomedical literature search engines where queries are generally short and top documents account for most of the total clicks. Motivated by this, we present a novel architecture for biomedical literature search, namely Log-Augmented DEnse Retrieval (LADER), which is a simple plug-in module that augments a dense retriever with the click logs retrieved from similar training queries. Specifically, LADER finds both similar documents and queries to the given query by a dense retriever. Then, LADER scores relevant (clicked) documents of similar queries weighted by their similarity to the input query. The final document scores by LADER are the average of (1) the document similarity scores from the dense retriever and (2) the aggregated document scores from the click logs of similar queries. Despite its simplicity, LADER achieves new state-of-the-art (SOTA) performance on TripClick, a recently released benchmark for biomedical literature retrieval. On the frequent (HEAD) queries, LADER largely outperforms the best retrieval model by 39% relative NDCG@10 (0.338 v.s. 0.243). LADER also achieves better performance on the less frequent (TORSO) queries with 11% relative NDCG@10 improvement over the previous SOTA (0.303 v.s. 0.272). On the rare (TAIL) queries where similar queries are scarce, LADER still compares favorably to the previous SOTA method (NDCG@10: 0.310 v.s. 0.295). On all queries, LADER can improve the performance of a dense retriever by 24%-37% relative NDCG@10 while not requiring additional training, and further performance improvement is expected from more logs. Our regression analysis has shown that queries that are more frequent, have higher entropy of query similarity and lower entropy of document similarity, tend to benefit more from log augmentation.
Paper Structure (18 sections, 10 equations, 4 figures, 2 tables)

This paper contains 18 sections, 10 equations, 4 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Methods
  3. Dense Retriever Training
  4. LADER Inference
  5. Experiments
  6. Settings
  7. Dataset.
  8. Configuration.
  9. Baseline methods for comparison.
  10. Main Results
  11. Comparison with benchmark baselines
  12. Comparison with BERT DOT
  13. Improvement over the backbone dense retriever
  14. Analysis
  15. Importance of trained dense retrievers.
  16. ...and 3 more sections

Figures (4)

  • Figure 1: An example of using click logs of retrieved similar queries to augment biomedical literature search.
  • Figure 2: Overall architecture of the LADER model.
  • Figure 3: NDCG@10 of LADER with different proportions of training queries to retrieve in log-augmentation.
  • Figure 4: Feature coefficients in the regression analysis. Positive values indicate more gains from log augmentation.