A Weakly Supervised Approach to Evaluating Single-Document Summarization via Negative Sampling

Anonymous

Abstract: 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 via simple negative sampling methods. In cross-domain tests, our strategy outperforms baselines with promising improvements, and show a great advantage in gauging linguistic qualities over all metrics. We hope this study can inspire more research using similar strategies. Our code is at https://anonymous.4open.science/r/37CF.

Paper Type: short