English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Acronyms and long-forms are commonly found in research documents, more so indocuments from scientific and legal domains. Many acronyms used in suchdocuments are domain-specific and are very rarely found in normal text corpora.Owing to this, transformer-based NLP models often detect OOV (Out ofVocabulary) for acronym tokens, especially for non-English languages, and theirperformance suffers while linking acronyms to their long forms duringextraction. Moreover, pretrained transformer models like BERT are notspecialized to handle scientific and legal documents. With these points beingthe overarching motivation behind this work, we propose a novel frameworkCABACE: Character-Aware BERT for ACronym Extraction, which takes into accountcharacter sequences in text and is adapted to scientific and legal domains bymasked language modelling. We further use an objective with an augmented lossfunction, adding the max loss and mask loss terms to the standard cross-entropyloss for training CABACE. We further leverage pseudo labelling and adversarialdata generation to improve the generalizability of the framework. Experimentalresults prove the superiority of the proposed framework in comparison tovarious baselines. Additionally, we show that the proposed framework is bettersuited than baseline models for zero-shot generalization to non-Englishlanguages, thus reinforcing the effectiveness of our approach. Our teamBacKGProp secured the highest scores on the French dataset, second-highest onDanish and Vietnamese, and third-highest in the English-Legal dataset on theglobal leaderboard for the acronym extraction (AE) shared task at SDU AAAI-22.

CABACE: Injecting Character Sequence Information and Domain Knowledge for Enhanced Acronym and Long-Form Extraction