A condense-then-select strategy for text summarization

Select-then-compress is a popular hybrid, framework for text summarization due to its high efficiency. This first selects salient sentences and then independently condenses each of the selected into concise version. However, compressing separately ignores context information document, therefore prone delete information. To address this limitation, we propose novel condense-then-select summarization. Our concurrently document sentence. Original their compressed versions become candidates extraction. Finally, an extractor utilizes select assembles them summary. If deleted during condensing, can original sentence retain Thus, our helps avoid loss information, while preserving efficiency sentence-level compression. Experiment results1 on CNN/DailyMail, DUC-2002, Pubmed datasets demonstrate that outperforms select-then-compress other strong baselines.