E2E NLG Challenge: Neural Models vs. Templates

WS 2018 Yevgeniy PuzikovIryna Gurevych

E2E NLG Challenge is a shared task on generating restaurant descriptions from sets of key-value pairs. This paper describes the results of our participation in the challenge... (read more)

PDF Abstract

Evaluation Results from the Paper


TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK COMPARE
Data-to-Text Generation E2E NLG Challenge TUDA BLEU 56.57 # 9
Data-to-Text Generation E2E NLG Challenge TUDA NIST 7.4544 # 9
Data-to-Text Generation E2E NLG Challenge TUDA METEOR 45.29 # 1
Data-to-Text Generation E2E NLG Challenge TUDA ROUGE-L 66.14 # 9
Data-to-Text Generation E2E NLG Challenge TUDA CIDEr 1.8206 # 8