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

Nowadays on E-commerce platforms, products are presented to the customerswith multiple modalities. These multiple modalities are significant for aretrieval system while providing attracted products for customers. Therefore,how to take into account those multiple modalities simultaneously to boost theretrieval performance is crucial. This problem is a huge challenge to us due tothe following reasons: (1) the way of extracting patch features with thepre-trained image model (e.g., CNN-based model) has much inductive bias. It isdifficult to capture the efficient information from the product image inE-commerce. (2) The heterogeneity of multimodal data makes it challenging toconstruct the representations of query text and product including title andimage in a common subspace. We propose a novel Adversarial Cross-modal EnhancedBERT (ACE-BERT) for efficient E-commerce retrieval. In detail, ACE-BERTleverages the patch features and pixel features as image representation. Thusthe Transformer architecture can be applied directly to the raw imagesequences. With the pre-trained enhanced BERT as the backbone network, ACE-BERTfurther adopts adversarial learning by adding a domain classifier to ensure thedistribution consistency of different modality representations for the purposeof narrowing down the representation gap between query and product.Experimental results demonstrate that ACE-BERT outperforms the state-of-the-artapproaches on the retrieval task. It is remarkable that ACE-BERT has alreadybeen deployed in our E-commerce's search engine, leading to 1.46% increase inrevenue.

ACE-BERT: Adversarial Cross-modal Enhanced BERT for E-commerce Retrieval