Abstract
In this paper, we introduce the IMPortance-awaReness maskIng NeTwork (IMPRINT), a novel approach to enhance the robustness of document retrieval systems against query variations, particularly those containing mis-spellings. Unlike previous models that treat all query components (words/features) equally, IMPRINT prioritizes the most important components while masking out less relevant ones. Specifically, we propose a Mutual Information-based measure to quantify component importance and integrate it into a dynamic masking mechanism that adjusts the retention probability of each component. Our method is evaluated on a combination of three benchmark datasets and three types of query variations. The experimental results show substantial performance gains compared to state-of-the-art models, achieving an average improvement of 1.2 absolute MRR@10 points in retrieval accuracy.
| Original language | English |
|---|---|
| Title of host publication | Proceedings of 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2025), April 6 - April 11, 2025, Hyderabad, India |
| Editors | Bhaskar D Rao, Isabel Trancoso, Gaurav Sharma, Neelesh B. Mehta |
| Place of Publication | U.S. |
| Publisher | IEEE |
| Number of pages | 5 |
| ISBN (Electronic) | 9798350368741 |
| ISBN (Print) | 9798350368741 |
| DOIs | |
| Publication status | Published - 2025 |
| Event | ICASSP (Conference) - Hyderabad, India Duration: 6 Apr 2025 → 11 Apr 2025 |
Publication series
| Name | ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings |
|---|---|
| ISSN (Print) | 1520-6149 |
Conference
| Conference | ICASSP (Conference) |
|---|---|
| Country/Territory | India |
| City | Hyderabad |
| Period | 6/04/25 → 11/04/25 |
Bibliographical note
Publisher Copyright:© 2025 IEEE.
Keywords
- Document Retrieval
- Model Robustness
- Query Variations
- Word/Feature Importance