Scholarlink

Context <p>Empirical studies in software engineering mainly rely on the data available on code-hosting platforms, being <span>GitHub</span> the most representative. Nevertheless, in the last years, the emergence of Machine Learning (ML) has led to the development of platforms specifically designed for hosting ML-based projects, with <span>Hugging Face Hub</span> (HFH) as the most popular one. So far, there have been no studies evaluating the potential of HFH for such studies.</p> Objective <p>We aim at performing an exploratory study of the current state of HFH and its suitability to be used as a source platform for empirical studies.</p> Method <p>We conduct a qualitative and quantitative analysis of HFH. The former will be performed by comparing the features of HFH with those of other code-hosting platforms, such as <span>GitHub</span> and <span>GitLab</span>. The latter will be performed by analyzing the data available in HFH.</p> Results <p>We propose a feature framework to characterize HFH and report on the current usage of the platform, both in terms of number and types of projects (and surrounding community) and the features they mostly rely on.</p> Conclusions <p>The results confirm that HFH offers enough features and diverse enough data to be the source of relevant empirical studies on the development, evolution and usage of AI-related projects. The results also triggered a discussion on aspects of HFH that should be considered when performing such empirical studies.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

On the suitability of hugging face hub for empirical studies

  • 期刊论文

  • 发表日期：

    2025年1月18日

查看全文
  • Adem Ait,
  • Javier Luis Cánovas Izquierdo,
  • Jordi Cabot

摘要

Context

Empirical studies in software engineering mainly rely on the data available on code-hosting platforms, being GitHub the most representative. Nevertheless, in the last years, the emergence of Machine Learning (ML) has led to the development of platforms specifically designed for hosting ML-based projects, with Hugging Face Hub (HFH) as the most popular one. So far, there have been no studies evaluating the potential of HFH for such studies.

Objective

We aim at performing an exploratory study of the current state of HFH and its suitability to be used as a source platform for empirical studies.

Method

We conduct a qualitative and quantitative analysis of HFH. The former will be performed by comparing the features of HFH with those of other code-hosting platforms, such as GitHub and GitLab. The latter will be performed by analyzing the data available in HFH.

Results

We propose a feature framework to characterize HFH and report on the current usage of the platform, both in terms of number and types of projects (and surrounding community) and the features they mostly rely on.

Conclusions

The results confirm that HFH offers enough features and diverse enough data to be the source of relevant empirical studies on the development, evolution and usage of AI-related projects. The results also triggered a discussion on aspects of HFH that should be considered when performing such empirical studies.