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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Multi-label text classification (MLC) is a challenging task in settings oflarge label sets, where label support follows a Zipfian distribution. In thispaper, we address this problem through retrieval augmentation, aiming toimprove the sample efficiency of classification models. Our approach closelyfollows the standard MLC architecture of a Transformer-based encoder pairedwith a set of classification heads. In our case, however, the input documentrepresentation is augmented through cross-attention to similar documentsretrieved from the training set and represented in a task-specific manner. Weevaluate this approach on four datasets from the legal and biomedical domains,all of which feature highly skewed label distributions. Our experiments showthat retrieval augmentation substantially improves model performance on thelong tail of infrequent labels especially so for lower-resource trainingscenarios and more challenging long-document data scenarios.

Retrieval-augmented Multi-label Text Classification