Of Human Criteria and Automatic Metrics: A Benchmark of the Evaluation of Story Generation - Archive ouverte HAL Access content directly
Conference Papers Year :

Of Human Criteria and Automatic Metrics: A Benchmark of the Evaluation of Story Generation

(1, 2) , (3, 4) , (5, 6, 2) , (7, 8, 2)
1
2
3
4
5
6
7
8

Abstract

Research on Automatic Story Generation (ASG) relies heavily on human and automatic evaluation. However, there is no consensus on which human evaluation criteria to use, and no analysis of how well automatic criteria correlate with them. In this paper, we propose to re-evaluate ASG evaluation. We introduce a set of 6 orthogonal and comprehensive human criteria, carefully motivated by the social sciences literature. We also present HANNA, an annotated dataset of 1,056 stories produced by 10 different ASG systems. HANNA allows us to quantitatively evaluate the correlations of 72 automatic metrics with human criteria. Our analysis highlights the weaknesses of current metrics for ASG and allows us to formulate practical recommendations for ASG evaluation.
Fichier principal
Vignette du fichier
2208.11646.pdf (11.11 Mo) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-03801053 , version 1 (06-10-2022)

Identifiers

  • HAL Id : hal-03801053 , version 1

Cite

Cyril Chhun, Pierre Colombo, Fabian M Suchanek, Chloé Clavel. Of Human Criteria and Automatic Metrics: A Benchmark of the Evaluation of Story Generation. 29th International Conference on Computational Linguistics (COLING 2022), Oct 2022, Gyeongju, South Korea. ⟨hal-03801053⟩
1 View
1 Download

Share

Gmail Facebook Twitter LinkedIn More