We describe the University of Edinburgh’s submission to the First Shared Task on Automatic Minuting. We developed an English-language minuting system for Task A that combines BERT-based extractive summarization with logistic regression-based filtering and rule-based pre- and post-processing steps. In the human evaluation, our system averaged scores of 2.1 on adequacy, 3.9 on grammatical correctness, and 3.3 on fluency.
Cite as: Williams, P., Haddow, B. (2021) Team UEDIN @ AutoMin 2021: Creating Minutes by Learning to Filter an Extracted Summary. Proc. First Shared Task on Automatic Minuting at Interspeech 2021, 78-81, doi: 10.21437/AutoMin.2021-10
@inproceedings{williams21_automin,
author={Philip Williams and Barry Haddow},
title={{Team UEDIN @ AutoMin 2021: Creating Minutes by Learning to Filter an Extracted Summary}},
year=2021,
booktitle={Proc. First Shared Task on Automatic Minuting at Interspeech 2021},
pages={78--81},
doi={10.21437/AutoMin.2021-10}
}