Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Wullschleger, Pascal, Zarharan, Majid, Daly, Donnacha, Pouly, Marc, Foster, Jennifer
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2505.11470
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866909982438457344
author Wullschleger, Pascal
Zarharan, Majid
Daly, Donnacha
Pouly, Marc
Foster, Jennifer
author_facet Wullschleger, Pascal
Zarharan, Majid
Daly, Donnacha
Pouly, Marc
Foster, Jennifer
contents We introduce two reference-free metrics for quality evaluation of taxonomies in the absence of labels. The first metric evaluates robustness by calculating the correlation between semantic and taxonomic similarity, addressing error types not considered by existing metrics. The second uses Natural Language Inference to assess logical adequacy. Both metrics are tested on five taxonomies and are shown to correlate well with F1 against ground truth taxonomies. We further demonstrate that our metrics can predict downstream performance in hierarchical classification when used with label hierarchies.
format Preprint
id arxiv_https___arxiv_org_abs_2505_11470
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Reference-Free Evaluation of Taxonomies
Wullschleger, Pascal
Zarharan, Majid
Daly, Donnacha
Pouly, Marc
Foster, Jennifer
Computation and Language
We introduce two reference-free metrics for quality evaluation of taxonomies in the absence of labels. The first metric evaluates robustness by calculating the correlation between semantic and taxonomic similarity, addressing error types not considered by existing metrics. The second uses Natural Language Inference to assess logical adequacy. Both metrics are tested on five taxonomies and are shown to correlate well with F1 against ground truth taxonomies. We further demonstrate that our metrics can predict downstream performance in hierarchical classification when used with label hierarchies.
title Reference-Free Evaluation of Taxonomies
topic Computation and Language
url https://arxiv.org/abs/2505.11470