Data To Text Generation On E2E Nlg Challenge

Métriques

BLEU

CIDEr

METEOR

NIST

ROUGE-L

Résultats

Résultats de performance de divers modèles sur ce benchmark

						Paper Title
S_1^R	68.60	2.37	45.25	8.73	70.82	Pragmatically Informative Text Generation
EDA_CS	67.05	2.2355	44.49	8.5150	68.94	Copy mechanism and tailored training for character-based data-to-text generation
TrICy (trK = 0)	66.43	-	-	-	70.14	TrICy: Trigger-guided Data-to-text Generation with Intent aware Attention-Copy
Slug	66.19	-	44.54	8.6130	67.72	A Deep Ensemble Model with Slot Alignment for Sequence-to-Sequence Natural Language Generation
TGen	65.93	2.2338	44.83	8.6094	68.50	Findings of the E2E NLG Challenge
EDA_CS (TL)	65.80	2.1803	45.16	8.5615	67.40	Copy mechanism and tailored training for character-based data-to-text generation
Sys1-Primary	65.61	2.2183	45.17	8.5105	68.39	TNT-NLG, System 1: Using a statistical NLG to massively augment crowd-sourced data for neural generation
Zhang	65.45	2.1012	43.92	8.1804	70.83	Attention Regularized Sequence-to-Sequence Learning for E2E NLG Challenge
Self-memory	65.11	2.16	46.11	8.35	68.41	Self-training from Self-memory in Data-to-text Generation
Gong	64.22	2.2721	44.69	8.3453	66.45	Technical Report for E2E NLG Challenge
TUDA	56.57	1.8206	45.29	7.4544	66.14	E2E NLG Challenge: Neural Models vs. Templates

0 of 11 row(s) selected.

Data To Text Generation On E2E Nlg Challenge | SOTA | HyperAI