jobGBERT Open-Source Pre-trained Language Model - Empowering Text Processing in German Recruitment Ad Domain

Home

Jobgbert

Developed by agne

A pre-trained language model based on Transformer architecture, optimized for German job advertisements

Large Language Model

Transformers

German#German Job Ad Optimization #Swiss Job Market #BERT Domain Adaptation

Downloads 41

Release Time : 6/3/2022

Model Overview

This model is built upon deepset/gbert-base and achieves domain adaptation through continuous pre-training on 4 million Swiss German job advertisements from 1990-2020, primarily used for German job advertisement analysis.

Model Features

Domain Adaptation

Specifically optimized for German job advertisements through continuous pre-training on 4 million Swiss German job ads

Based on gbert-base

Built upon the high-quality German BERT model deepset/gbert-base

Academic Research Support

Developed with support from the University of Zurich's Department of Sociology, ensuring academic rigor

Model Capabilities

German text understanding

Job advertisement analysis

Domain-specific language modeling

Use Cases

Job Advertisement Analysis

Job Requirement Analysis

Analyze job requirements and skill descriptions in advertisements

Can identify key skills and qualification requirements in job ads

Market Trend Monitoring

Track changes in employment market trends

Can be used to monitor demand changes in specific industries or occupations

Social Science Research

Labor Market Research

Support labor market research in sociology

Helps researchers analyze job market structures and changes

🚀 jobGBERT

jobGBERT is a domain - adapted language model based on transformers, specifically designed for German - speaking job advertisements. It builds on the [deepset/gbert - base](https://huggingface.co/deepset/gbert - base) model and is adapted to the job advertisement domain through continued in - domain pretraining on 4 million German - speaking job ads from Switzerland between 1990 and 2020 (5.9 GB of data).

🚀 Quick Start

The model can be used for masked language modeling, but it is primarily intended to be fine - tuned on a downstream task.

✨ Features

Domain Adaptation: Adapted to the domain of German - speaking job advertisements.
Based on a Strong Foundation: Built upon the [deepset/gbert - base](https://huggingface.co/deepset/gbert - base) model.

📚 Documentation

Overview

Property	Details
Model Type	BERT base
Language	German
Domain	Job advertisements
Related Model	[agne/jobBERT - de](https://huggingface.co/agne/jobBERT - de)

Intended usage and limitations

⚠️ Important Note

The model inherits potential bias from its base model and may contain biases and stereotypes common in job advertisements. It is trained on German - speaking job ads from Switzerland.

You can use the model for masked language modeling, but it's intended to be fine - tuned on a downstream task.

License

The model is released under the Creative Commons Attribution - NonCommercial - ShareAlike 4.0 International License (cc - by - nc - sa - 4.0).

Please use the following citation when using our model:

@inproceedings{
    title = "Evaluation of Transfer Learning and Domain Adaptation for Analyzing German-Speaking Job Advertisements",
    author = "Gnehm, Ann-Sophie and
      Bühlmann, Eva and
      Clematide, Simon",
    booktitle = "Proceedings of the 13th Language Resources and Evaluation Conference",
    month = june,
    year = "2022",
    address = "Marseille, France",
    publisher = "European Language Resources Association",
}

About us

Ann - Sophie Gnehm: gnehm [at] soziologie.uzh.ch
Eva Bühlmann: bühlmann [at] soziologie.uzh.ch
Simon Clematide: simon.clematide [at] cl.uzh.ch

The Swiss Job Market Monitor aims at systematically expanding scientific knowledge about the job market and improving labour market transparency by informing the general public about current developments on the job market.

Get in touch: Mail Website Zenodo SWISSUbase

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご