EUVIMEDEUVIMED entdecken
Open Access · CC BY-SA 4.0

Efficient Neural Ranking Using Forward Indexes and Lightweight Encoders

Jurek Leonhardt, Henrik Müller, Koustav Rudra, Megha Khosla, Abhijit Anand, Avishek Anand

ACM Transactions on Information Systems · 30.09.2024 · 42 (5) · 1-34

DOI: 10.1145/3631939 ↗

Kurz gefasst

Fachartikel in der Zeitschrift ACM Transactions on Information Systems, online erschienen am 29.04.2024. Verfasst von Jurek Leonhardt und 5 weiteren Autor:innen. Die Arbeit wurde bisher 7-mal zitiert (Crossref) und nennt 92 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-SA 4.0). Gefördert von: European Union — Horizon 2020 Program under the scheme, Integrating Activities for Advanced Communities, SoBigData++: European Integrated Infrastructure for Social Mining and Big Data Analytics.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Dual-encoder-based dense retrieval models have become the standard in IR. They employ large Transformer-based language models, which are notoriously inefficient in terms of resources and latency. We propose Fast-Forward indexes—vector forward indexes which exploit the semantic matching capabilities of dual-encoder models for efficient and effective re-ranking. Our framework enables re-ranking at very high retrieval depths and combines the merits of both lexical and semantic matching via score interpolation. Furthermore, in order to mitigate the limitations of dual-encoders, we tackle two main challenges: Firstly, we improve computational efficiency by either pre-computing representations, avoiding unnecessary computations altogether, or reducing the complexity of encoders. This allows us to considerably improve ranking efficiency and latency. Secondly, we optimize the memory footprint and maintenance cost of indexes; we propose two complementary techniques to reduce the index size and show that, by dynamically dropping irrelevant document tokens, the index maintenance efficiency can be improved substantially. We perform an evaluation to show the effectiveness and efficiency of Fast-Forward indexes—our method has low latency and achieves competitive results without the need for hardware acceleration, such as GPUs.

Förderung

  • European Union — Horizon 2020 Program under the scheme
  • Integrating Activities for Advanced Communities
  • SoBigData++: European Integrated Infrastructure for Social Mining and Big Data Analytics
  • Science and Engineering Research Board, Department of Science and Technology, Government of India
  • Microsoft Academic Partnership Grant 2023
  • DST-INSPIRE Faculty Fellowship

Zitieren

APA

Leonhardt, J., Müller, H., Rudra, K., Khosla, M., Anand, A., & Anand, A. (2024). Efficient Neural Ranking Using Forward Indexes and Lightweight Encoders. ACM Transactions on Information Systems, 42(5), 1-34. https://doi.org/10.1145/3631939

Vancouver

Leonhardt J, Müller H, Rudra K, Khosla M, Anand A, Anand A. Efficient Neural Ranking Using Forward Indexes and Lightweight Encoders. ACM Transactions on Information Systems. 2024;42(5):1-34. doi:10.1145/3631939

BibTeX
@article{Leonhardt2024,
  author = {Leonhardt, Jurek and Müller, Henrik and Rudra, Koustav and Khosla, Megha and Anand, Abhijit and Anand, Avishek},
  title = {Efficient Neural Ranking Using Forward Indexes and Lightweight Encoders},
  journal = {ACM Transactions on Information Systems},
  year = {2024},
  volume = {42},
  number = {5},
  pages = {1--34},
  doi = {10.1145/3631939}
}
Kostenlos starten →