English

Bipartite Graph Pre-training for Unsupervised Extractive Summarization with Graph Convolutional Auto-Encoders

Computation and Language 2023-10-31 v1 Artificial Intelligence

Abstract

Pre-trained sentence representations are crucial for identifying significant sentences in unsupervised document extractive summarization. However, the traditional two-step paradigm of pre-training and sentence-ranking, creates a gap due to differing optimization objectives. To address this issue, we argue that utilizing pre-trained embeddings derived from a process specifically designed to optimize cohensive and distinctive sentence representations helps rank significant sentences. To do so, we propose a novel graph pre-training auto-encoder to obtain sentence embeddings by explicitly modelling intra-sentential distinctive features and inter-sentential cohesive features through sentence-word bipartite graphs. These pre-trained sentence representations are then utilized in a graph-based ranking algorithm for unsupervised summarization. Our method produces predominant performance for unsupervised summarization frameworks by providing summary-worthy sentence representations. It surpasses heavy BERT- or RoBERTa-based sentence representations in downstream tasks.

Keywords

Cite

@article{arxiv.2310.18992,
  title  = {Bipartite Graph Pre-training for Unsupervised Extractive Summarization with Graph Convolutional Auto-Encoders},
  author = {Qianren Mao and Shaobo Zhao and Jiarui Li and Xiaolei Gu and Shizhu He and Bo Li and Jianxin Li},
  journal= {arXiv preprint arXiv:2310.18992},
  year   = {2023}
}

Comments

Accepted by the 2023 Conference on Empirical Methods in Natural Language Processing (EMNLP 2023)

R2 v1 2026-06-28T13:05:03.549Z