set-encoder-novelty-base Open Source Model - Efficient Paragraph Reordering, Suitable for Novelty-aware Tasks

Set Encoder Novelty Base

Developed by webis

Set-Encoder is a cross-encoder architecture specifically designed for efficient and permutation-invariant paragraph reordering, particularly suitable for novelty-aware re-ranking tasks.

Text Embedding

Safetensors

Open Source License:Apache-2.0 #Paragraph Reordering #Permutation Invariance #Cross-Encoder

Downloads 14

Release Time : 4/3/2025

Model Overview

This model is based on the ELECTRA architecture, achieving listwise paragraph reordering through cross-encoder and employing permutation-invariant inter-paragraph attention mechanisms, focusing on improving re-ranking performance in information retrieval.

Model Features

Permutation Invariance

The model employs a special attention mechanism, ensuring paragraph reordering results are unaffected by input sequence.

Novelty Awareness

Specifically optimized for novelty requirements in information retrieval, effectively identifying and promoting rankings of novel content.

Efficient Reordering

As a cross-encoder architecture, it achieves efficient paragraph reordering while maintaining high performance.

Model Capabilities

Paragraph Reordering

Information Retrieval Optimization

Novelty Detection

Use Cases

Information Retrieval

Search Engine Result Reordering

Optimizes the initial results returned by search engines to enhance novelty and relevance.

Significantly improves nDCG@10 metrics on TREC DL dataset.

QA System Paragraph Ranking

Optimizes ranking of candidate answer paragraphs retrieved in QA systems.

🚀 Set-Encoder: Permutation-Invariant Inter-Passage Attention for Listwise Passage Re-Ranking with Cross-Encoders

This model is a cross-encoder architecture designed for efficient and permutation-invariant passage re-ranking.

🚀 Quick Start

This model is presented in the paper Set-Encoder: Permutation-Invariant Inter-Passage Attention for Listwise Passage Re-Ranking with Cross-Encoders. It's designed for efficient and permutation - invariant passage re - ranking.

Code: https://github.com/webis-de/set-encoder

We offer the following fine - tuned models for novelty - aware re - ranking.

To reproduce the results, run the following command using the Lightning IR library and the configuration files from the repository linked above:

lightning-ir re_rank --config ./configs/re-rank-novelty.yaml --model.model_name_or_path <MODEL_NAME>

(alpha nDCG@10, alpha = 0.99 on TREC DL 19 and TREC DL 20)

Model Name	TREC DL 19 (BM25)	TREC DL 20 (BM25)	TREC DL 19 (ColBERTv2)	TREC DL 20 (ColBERTv2)
webis/set-encoder-novelty-base	0.805	0.721	0.821	0.803

📦 Installation

No specific installation steps are provided in the original document.

📚 Documentation

Model Information

Property	Details
Model Type	Cross - encoder
Base Model	google/electra - base - discriminator
Pipeline Tag	text - ranking
Library Name	lightning - ir
License	Apache - 2.0

📄 License

This project is licensed under the Apache - 2.0 license.

📜 Citation

If you use this code or the models in your research, please cite our paper:

@InProceedings{schlatt:2025,
  address =                  {Berlin Heidelberg New York},
  author =                   {Ferdinand Schlatt and Maik Fr{\"o}be and Harrisen Scells and Shengyao Zhuang and Bevan Koopman and Guido Zuccon and Benno Stein and Martin Potthast and Matthias Hagen},
  booktitle =                {Advances in Information Retrieval. 47th European Conference on IR Research (ECIR 2025)},
  doi =                      {10.1007/978-3-031-88711-6_1},
  month =                    apr,
  publisher =                {Springer},
  series =                   {Lecture Notes in Computer Science},
  site =                     {Lucca, Italy},
  title =                    {{Set-Encoder: Permutation-Invariant Inter-Passage Attention for Listwise Passage Re-Ranking with Cross-Encoders}},
  year =                     2025
}

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご