Papers by Jesse Egbert
Beyond the English Web: Zero-Shot Cross-Lingual and Lightweight Monolingual Classification of Registers (2021.eacl-srw)
Copied to clipboard
Liina Repo, Valtteri Skantsi, Samuel Rönnqvist, Saara Hellström, Miika Oinonen, Anna Salmela, Douglas Biber, Jesse Egbert, Sampo Pyysalo, Veronika Laippala
| Challenge: | Existing studies on register classification for web documents have limited results due to skewed datasets and low performance. |
| Approach: | They propose two new register-annotated corpora for French and Swedish . they show that deep pre-trained language models perform strongly in these languages . |
| Outcome: | The proposed models outperform existing models in English and Finnish and can match or surpass existing models. |
Jump-Starting Item Parameters for Adaptive Language Tests (2021.emnlp-main)
Copied to clipboard
| Challenge: | Prior work has addressed ‘cold start’ estimation of item difficulties without piloting, but a multi-task generalized linear model with BERT features is needed to jump-start new items without pilot. |
| Approach: | They propose a multi-task generalized linear model with BERT features to jump-start new item difficulties without piloting them first. |
| Outcome: | The proposed model compares test-taker proficiency, item difficulty, and language proficiency frameworks like the Common European Framework of Reference (CEFR). |