skill-classifier-base (Legacy)

A newer, improved model is available!
We highly recommend using loyoladatamining/skill-classifier-base-v2 instead of this model.

Model Description

This is a legacy BERT-based model fine-tuned for sentence-level binary classification of skill statements. It was built on top of prajjwal1/bert-small.

Why Upgrade?

When evaluated on the loyoladatamining/usajobs_validation dataset, the newer version demonstrates higher performance:

Model	Accuracy	F-1
skill-classifier-base (This Model)	0.8335	0.8437
skill-classifier-base-v2 (Recommended)	0.9748	0.9749

Citation

If you do find this model useful in your work, please consider citing:

@article{meisenbacher2025extracting,
  title={Extracting O* NET Features from the NLx Corpus to Build Public Use Aggregate Labor Market Data},
  author={Meisenbacher, Stephen and Nestorov, Svetlozar and Norlander, Peter},
  year={2025}
}