Papers › Pre-training Meets Clustering: A Hybrid Extractive Multi-document Summarization Model

Pre-training Meets Clustering: A Hybrid Extractive Multi-document Summarization Model

25 May 2023International Conference on Hybrid Intelligent Systems 2023 5archive 2025-07-28

Akanksha Karotia, Seba Susan

In this era where a large amount of information has flooded the Internet, manual extraction and consumption of relevant information is very difficult and time-consuming. Therefore, an automated document summarization tool is necessary to excerpt important information from a set of documents that have similar or related subjects. Multi-document summarization allows retrieval of important and relevant content from multiple documents while minimizing redundancy. A multi-document text summarization system is developed in this study using an unsupervised extractive-based approach. The proposed model is a fusion of two learning paradigms: the T5 pre-trained transformer model and the K-Means clustering algorithm. We perform the experiments on the benchmark news article corpus Document Understanding Conference (DUC2004). The ROUGE evaluation metrics were used to estimate the performance of the proposed approach on the DUC2004. Outcomes validate that our proposed model shows greatly enhanced performance as compared to the existent unsupervised state-of-the-art approaches.

Paper PDF Code

Code

Akankshakarotia/Pre-training-meets-Clustering-A-Hybrid-Extractive-Multi-Document-Summarization-Model official report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Clustering Document Summarization Extractive Text Summarization Multi-Document Summarization Retrieval Text Summarization Unsupervised Text Summarization document understanding

Results from the paper archive 2025-07-28

Task	Dataset	Model	Metric	Value	Rank at snapshot	Leaderboard	Report
Extractive Text Summarization	DUC 2004	Pre-training-meets-Clustering-A-Hybrid-Extractive-Multi-Document-Summarization-Model	Test ROGUE-1	34.013	#1 of 1	Archive leaderboard	report
Extractive Text Summarization	DUC 2004	Pre-training-meets-Clustering-A-Hybrid-Extractive-Multi-Document-Summarization-Model	Test ROGUE-2	8.266	#1 of 1	Archive leaderboard	report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

Adafactor Attention Attention Dropout BPE Dense Connections Dropout Gated Linear Unit Inverse Square Root Schedule Layer Normalization Linear Layer Multi-Head Attention Residual Connection SentencePiece Softmax T5 k-Means Clustering

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