Data for "SpaceTransformers: language modeling for space systems"

This dataset contains the labelled concepts used to generate the Concept Recognition results in "SpaceTransformers: Language Modeling for Space Systems" published by IEEE Access (10.1109/ACCESS.2021.3115659). In this paper, a novel family of domain-specific models, SpaceTransformers was further pre-trained from BERT, RoBERTa and SciBERT on unstructured data related to space systems. The models were then fine-tuned on a domain-specific Concept Recognition task.

Additional Links:
1) The code is available at github.com/strath-ace/smart-nlp.
2) The further pre-training corpus is available at doi.org/10.15129/8e1c
3353-ccbe-4835-b4f9-bffd6b5e058b.
3) The ECSS requirements used as fine-tuning corpus can be accessed through https://ecss.nl/standards/downloads/doors-download/.

If re-using this dataset, please do cite our work:
@ARTICLE{9548078,
author={Berquand, Audrey and Darm, Paul and Riccardi, Annalisa},
journal={IEEE Access},
title={SpaceTransformers: Language Modeling for Space Systems},
year={2021},
volume={9},
pages={133111-133122},
doi={10.1109/ACCESS.2021.3115659}}