A survey on evaluation of summarization methods

•Manual assessment is not re-usable.•Re-use of the gold standard by non-participants is often problematic.•Overlap-based metrics are not suitable for full text comparison-based evaluation.•GRAD exceeds word-based metrics to distinguish between generated and human written summaries.•Overlap metrics a...

Ausführliche Beschreibung

Gespeichert in:

Bibliographische Detailangaben
Veröffentlicht in:	Information processing & management 2019-09, Vol.56 (5), p.1794-1814
Hauptverfasser:	Ermakova, Liana, Cossu, Jean Valère, Mothe, Josiane
Format:	Artikel
Sprache:	eng
Schlagworte:	Annotations Assessment metrics Automatic summarization Benchmarks Computer Science Evaluation campaigns Extraction Extractive summarization Feature extraction Humanities and Social Sciences Information processing Information retrieval Literature Natural language processing ROUGE Scientific papers Summaries Text compression
Online-Zugang:	Volltext
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!

Schreiben Sie den ersten Kommentar!