Skip to main content

Table 5 DRAGON-AI BERTScore evaluation

From: Dynamic Retrieval Augmented Generation of Ontologies using Artificial Intelligence (DRAGON-AI)

method

model name

F1

P

R

DRAGON

gpt-3.5-turbo

0.923

0.933

0.914

DRAGON

gpt-4

0.920

0.928

0.912

DRAGON

nous-hermes-13b

0.916

0.922

0.910

Random Selection

n/a

0.838

0.845

0.832

  1. Scoring of generated definition against curated definition using BERTScore