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

Deep Models are increasingly becoming prevalent in summarization problems(e.g. document, video and images) due to their ability to learn complex featureinteractions and representations. However, they do not model characteristicssuch as diversity, representation, and coverage, which are also very importantfor summarization tasks. On the other hand, submodular functions naturallymodel these characteristics because of their diminishing returns property. Mostapproaches for modelling and learning submodular functions rely on very simplemodels, such as weighted mixtures of submodular functions. Unfortunately, thesemodels only learn the relative importance of the different submodular functions(such as diversity, representation or importance), but cannot learn morecomplex feature representations, which are often required for state-of-the-artperformance. We propose Deep Submodular Networks (DSN), an end-to-end learningframework that facilitates the learning of more complex features and richerfunctions, crafted for better modelling of all aspects of summarization. TheDSN framework can be used to learn features appropriate for summarization fromscratch. We demonstrate the utility of DSNs on both generic and query focusedimage-collection summarization, and show significant improvement over thestate-of-the-art. In particular, we show that DSNs outperform simple mixturemodels using off the shelf features. Secondly, we also show that just usingfour submodular functions in a DSN with end-to-end learning performs comparablyto the state-of-the-art mixture model with a hand-crafted set of 594 componentsand outperforms other methods for image collection summarization.

Deep Submodular Networks for Extractive Data Summarization