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

The sparsity of labelled data is an obstacle to the development of RelationExtraction models and the completion of databases in various biomedical areas.While being of high interest in drug-discovery, the natural-productsliterature, reporting the identification of potential bioactive compounds fromorganisms, is a concrete example of such an overlooked topic. To mark the startof this new task, we created the first curated evaluation dataset and extractedliterature items from the LOTUS database to build training sets. To this end,we developed a new sampler inspired by diversity metrics in ecology, namedGreedy Maximum Entropy sampler, or GME-sampler(https://github.com/idiap/gme-sampler). The strategic optimization of bothbalance and diversity of the selected items in the evaluation set is importantgiven the resource-intensive nature of manual curation. After quantifying thenoise in the training set, in the form of discrepancies between the inputabstracts text and the expected output labels, we explored different strategiesaccordingly. Framing the task as an end-to-end Relation Extraction, weevaluated the performance of standard fine-tuning as a generative task andfew-shot learning with open Large Language Models (LLaMA 7B-65B). In additionto their evaluation in few-shot settings, we explore the potential of openLarge Language Models (Vicuna-13B) as synthetic data generator and propose anew workflow for this purpose. All evaluated models exhibited substantialimprovements when fine-tuned on synthetic abstracts rather than the originalnoisy data. We provide our best performing (f1-score=59.0) BioGPT-Large modelfor end-to-end RE of natural-products relationships along with all thegenerated synthetic data and the evaluation dataset. See more details athttps://github.com/idiap/abroad-re.

Relation Extraction in underexplored biomedical domains: A diversity-optimised sampling and synthetic data generation approach