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

In the biomedical domain, taxonomies organize the acquisition modalities ofscientific images in hierarchical structures. Such taxonomies leverage largesets of correct image labels and provide essential information about theimportance of a scientific publication, which could then be used in biocurationtasks. However, the hierarchical nature of the labels, the overhead ofprocessing images, the absence or incompleteness of labeled data, and theexpertise required to label this type of data impede the creation of usefuldatasets for biocuration. From a multi-year collaboration with biocurators andtext-mining researchers, we derive an iterative visual analytics and activelearning strategy to address these challenges. We implement this strategy in asystem called BI-LAVA Biocuration with Hierarchical Image Labeling throughActive Learning and Visual Analysis. BI-LAVA leverages a small set of imagelabels, a hierarchical set of image classifiers, and active learning to helpmodel builders deal with incomplete ground-truth labels, target a hierarchicaltaxonomy of image modalities, and classify a large pool of unlabeled images.BI-LAVA's front end uses custom encodings to represent data distributions,taxonomies, image projections, and neighborhoods of image thumbnails, whichhelp model builders explore an unfamiliar image dataset and taxonomy andcorrect and generate labels. An evaluation with machine learning practitionersshows that our mixed human-machine approach successfully supports domainexperts in understanding the characteristics of classes within the taxonomy, aswell as validating and improving data quality in labeled and unlabeledcollections.

BI-LAVA: Biocuration with Hierarchical Image Labeling through Active Learning and Visual Analysis