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

Adverse Event (ADE) extraction is one of the core tasks in digitalpharmacovigilance, especially when applied to informal texts. This task hasbeen addressed by the Natural Language Processing community using largepre-trained language models, such as BERT. Despite the great number ofTransformer-based architectures used in the literature, it is unclear which ofthem has better performances and why. Therefore, in this paper we perform anextensive evaluation and analysis of 19 Transformer-based models for ADEextraction on informal texts. We compare the performance of all the consideredmodels on two datasets with increasing levels of informality (forums posts andtweets). We also combine the purely Transformer-based models with twocommonly-used additional processing layers (CRF and LSTM), and analyze theireffect on the models performance. Furthermore, we use a well-establishedfeature importance technique (SHAP) to correlate the performance of the modelswith a set of features that describe them: model category (AutoEncoding,AutoRegressive, Text-to-Text), pretraining domain, training from scratch, andmodel size in number of parameters. At the end of our analyses, we identify alist of take-home messages that can be derived from the experimental data.

Extensive Evaluation of Transformer-based Architectures for Adverse Drug Events Extraction