Canonical automatic summary evaluation metrics, such as ROUGE, focus on lexical similarity which cannot well capture semantics nor linguistic quality and require a reference summary which is costly to obtain. Recently, there have been a growing number of efforts to alleviate either or both of the two drawbacks. In this paper, we present a proof-of-concept study to a weakly supervised summary evaluation approach without the presence of reference summaries. Massive data in existing summarization datasets are transformed for training by pairing documents with corrupted reference summaries. In cross-domain tests, our strategy outperforms baselines with promising improvements, and show a great advantage in gauging linguistic qualities over all metrics.
@article{arxiv.2005.06377,
title = {SueNes: A Weakly Supervised Approach to Evaluating Single-Document Summarization via Negative Sampling},
author = {Forrest Sheng Bao and Hebi Li and Ge Luo and Minghui Qiu and Yinfei Yang and Youbiao He and Cen Chen},
journal= {arXiv preprint arXiv:2005.06377},
year = {2022}
}