ROUGE metrics

GPTKB entity

Statements (40)
Predicate Object
gptkbp:instanceOf evaluation metric
gptkbp:abbreviation gptkb:Recall-Oriented_Understudy_for_Gisting_Evaluation
gptkbp:commonIn NLP research
text summarization competitions
gptkbp:creator gptkb:Chin-Yew_Lin
gptkbp:criticizedFor favoring extractive summaries
not capturing semantic similarity
gptkbp:hasVariant gptkb:ROUGE-L
ROUGE-1
ROUGE-2
ROUGE-SU4
https://www.w3.org/2000/01/rdf-schema#label ROUGE metrics
gptkbp:implementedIn gptkb:Java
gptkb:Perl
gptkb:Python
gptkbp:includes gptkb:ROUGE-L
gptkb:ROUGE-N
gptkb:ROUGE-S
gptkb:ROUGE-W
gptkbp:introducedIn 2004
gptkbp:measures precision
recall
F1-score
overlap between generated and reference texts
gptkbp:officialImplementation ROUGE-1.5.5.pl
gptkbp:publishedIn Proceedings of ACL 2004
gptkbp:relatedTo gptkb:BLEU_metric
gptkb:METEOR_metric
gptkbp:ROUGE-L longest common subsequence
gptkbp:ROUGE-N n-gram overlap
gptkbp:ROUGE-S skip-bigram overlap
gptkbp:ROUGE-W weighted longest common subsequence
gptkbp:usedBy gptkb:researchers
industry practitioners
gptkbp:usedFor automatic summarization evaluation
machine translation evaluation
gptkbp:bfsParent gptkb:ROUGE-N
gptkb:ROUGE-S
gptkb:ROUGE-W
gptkbp:bfsLayer 7