Scholarlink

We present insights about the use of different machine learning approaches from a project at the German Federal Employment Agency (FEA). The task at hand is the prediction of different entities and classes related to job advertisements. We adapted a gBERT-base model to our specific text domain of job advertisement texts. In line with earlier findings in the literature, we find the adapted model to outperform other approaches in some of our tasks. However, not all our tasks were able to profit from the domain adaptation.

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

Domain Adaptation of a BERT Model for Analyzing Job Advertisements at the German Federal Employment Agency

  • 章节

  • 出版模式：

    开放获取

  • 发表日期：

    2025年

查看全文
  • Lars Fiedler,
  • Barbara Hofmann,
  • Koen Loogman,
  • Tobias Scherl

摘要

We present insights about the use of different machine learning approaches from a project at the German Federal Employment Agency (FEA). The task at hand is the prediction of different entities and classes related to job advertisements. We adapted a gBERT-base model to our specific text domain of job advertisement texts. In line with earlier findings in the literature, we find the adapted model to outperform other approaches in some of our tasks. However, not all our tasks were able to profit from the domain adaptation.