Datasets › DUC 2004

DUC 2004

archive 2025-07-28

The DUC2004 dataset is a dataset for document summarization. Is designed and used for testing only. It consists of 500 news articles, each paired with four human written summaries. Specifically it consists of 50 clusters of Text REtrieval Conference (TREC) documents, from the following collections: AP newswire, 1998-2000; New York Times newswire, 1998-2000; Xinhua News Agency (English version), 1996-2000. Each cluster contained on average 10 documents.

Source: Discrete Optimization for Unsupervised Sentence Summarization with Word-Level Extraction Image Source: https://duc.nist.gov/duc2004/

Homepage Benchmarks Papers JSON

Benchmarks archive 2025-07-28

All 4 leaderboards whose dataset resolves to this page shown (sort by any header). "First row" is the archive's own first row at snapshot, in the archive's row order; nothing here re-ranks and metric direction is not asserted.

		First row (archive order)	Paper	Code
Text Summarization	DUC 2004 Task 1	Transformer+WDrop ROUGE-1 33.06	Rethinking Perturbations in Encoder-Decoders for Fast Training	takase/rethink_perturbations	13	Compare
Extractive Text Summarization	DUC 2004 Task 1	Abs ROUGE-1 26.55	A Neural Attention Model for Abstractive Sentence Summarization	tensorflow/models +3	1	Compare
Extractive Text Summarization	DUC 2004	Pre-training-meets-Clustering-A-Hybrid-Extractive-Multi-Document-Summarization-Model Test ROGUE-1 34.013	Pre-training Meets Clustering: A Hybrid Extractive...	Akankshakarotia/Pre-training-meets-Clustering-A-Hybrid-Extractive-Multi-Document-Summarization-Model	1	Compare
Multi-Document Summarization	DUC 2004	GCN: Personalized Discourse Graph ROUGE-1 38.23	Graph-based Neural Multi-Document Summarization	—	1	Compare

Papers archive 2025-07-28

14 shown of 14 papers with a leaderboard row on this dataset's benchmarks, newest first. The archive's own "papers using this dataset" list was never published, so this is the benchmark-backed subset; the archive's count for this dataset is 15. The Syntology column is from Syntology's graph (read 2026-09-24), stated per sample; it is not part of any archive number.

			Date	Samples run Syntology
Pre-training Meets Clustering: A Hybrid Extractive Multi-document Summarization Model	1	1	25 May 2023	not harvested
Rethinking Perturbations in Encoder-Decoders for Fast Training	1	1	5 Apr 2021	ran 1 of 1 samples (0 unverified; 1 pointer-only for licence)
All Word Embeddings from One Embedding	1	1	25 Apr 2020	ran 1 of 1 samples (0 unverified)
Sample Efficient Text Summarization Using a Single Pre-Trained Transformer	2	1	21 May 2019	not harvested
Positional Encoding to Control Output Sequence Length	1	1	16 Apr 2019	ran 0 of 4 samples (4 unverified)
Ensure the Correctness of the Summary: Incorporate Entailment Knowledge into Abstractive Sentence Summarization	0	1	1 Aug 2018	not harvested
A Reinforced Topic-Aware Convolutional Sequence-to-Sequence Model for Abstractive Text Summarization	0	1	9 May 2018	not harvested
Deep Recurrent Generative Decoder for Abstractive Text Summarization	1	1	2 Aug 2017	not harvested
Graph-based Neural Multi-Document Summarization	0	1	20 Jun 2017	not harvested
Selective Encoding for Abstractive Sentence Summarization	2	1	24 Apr 2017	not harvested
Cutting-off Redundant Repeating Generations for Neural Abstractive Summarization	0	1	31 Dec 2016	not harvested
Abstractive Sentence Summarization with Attentive Recurrent Neural Networks	0	1	1 Jun 2016	not harvested
Abstractive Text Summarization Using Sequence-to-Sequence RNNs and Beyond	4	1	19 Feb 2016	ran 5 of 6 samples (1 unverified)
A Neural Attention Model for Abstractive Sentence Summarization	4	3	2 Sep 2015	not harvested

Dataset loaders archive 2025-07-28

No loader listed in the archive.

Tasks archive 2025-07-28

Text Summarization Extractive Text Summarization Multi-Document Summarization

License archive 2025-07-28

No licence recorded in the archive. Absence here is not a statement about the dataset's terms.

Modalities archive 2025-07-28

Texts

Languages archive 2025-07-28

English

Variants archive 2025-07-28

DUC 2004 Task 1
DUC 2004

2 variant names, as the archive lists them.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections