Saved in:
Bibliographic Details
Main Authors: Toney-Wails, Autumn, Schoeberl, Christian, Dunham, James
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2403.09097
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913264564174848
author Toney-Wails, Autumn
Schoeberl, Christian
Dunham, James
author_facet Toney-Wails, Autumn
Schoeberl, Christian
Dunham, James
contents Identifying scientific publications that are within a dynamic field of research often requires costly annotation by subject-matter experts. Resources like widely-accepted classification criteria or field taxonomies are unavailable for a domain like artificial intelligence (AI), which spans emerging topics and technologies. We address these challenges by inferring a functional definition of AI research from existing expert labels, and then evaluating state-of-the-art chatbot models on the task of expert data annotation. Using the arXiv publication database as ground-truth, we experiment with prompt engineering for GPT chatbot models to identify an alternative, automated expert annotation pipeline that assigns AI labels with 94% accuracy. For comparison, we fine-tune SPECTER, a transformer language model pre-trained on scientific publications, that achieves 96% accuracy (only 2% higher than GPT) on classifying AI publications. Our results indicate that with effective prompt engineering, chatbots can be used as reliable data annotators even where subject-area expertise is required. To evaluate the utility of chatbot-annotated datasets on downstream classification tasks, we train a new classifier on GPT-labeled data and compare its performance to the arXiv-trained model. The classifier trained on GPT-labeled data outperforms the arXiv-trained model by nine percentage points, achieving 82% accuracy.
format Preprint
id arxiv_https___arxiv_org_abs_2403_09097
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle AI on AI: Exploring the Utility of GPT as an Expert Annotator of AI Publications
Toney-Wails, Autumn
Schoeberl, Christian
Dunham, James
Computation and Language
Identifying scientific publications that are within a dynamic field of research often requires costly annotation by subject-matter experts. Resources like widely-accepted classification criteria or field taxonomies are unavailable for a domain like artificial intelligence (AI), which spans emerging topics and technologies. We address these challenges by inferring a functional definition of AI research from existing expert labels, and then evaluating state-of-the-art chatbot models on the task of expert data annotation. Using the arXiv publication database as ground-truth, we experiment with prompt engineering for GPT chatbot models to identify an alternative, automated expert annotation pipeline that assigns AI labels with 94% accuracy. For comparison, we fine-tune SPECTER, a transformer language model pre-trained on scientific publications, that achieves 96% accuracy (only 2% higher than GPT) on classifying AI publications. Our results indicate that with effective prompt engineering, chatbots can be used as reliable data annotators even where subject-area expertise is required. To evaluate the utility of chatbot-annotated datasets on downstream classification tasks, we train a new classifier on GPT-labeled data and compare its performance to the arXiv-trained model. The classifier trained on GPT-labeled data outperforms the arXiv-trained model by nine percentage points, achieving 82% accuracy.
title AI on AI: Exploring the Utility of GPT as an Expert Annotator of AI Publications
topic Computation and Language
url https://arxiv.org/abs/2403.09097