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

With the prosperity of e-commerce industry, various modalities, e.g., visionand language, are utilized to describe product items. It is an enormouschallenge to understand such diversified data, especially via extracting theattribute-value pairs in text sequences with the aid of helpful image regions.Although a series of previous works have been dedicated to this task, thereremain seldomly investigated obstacles that hinder further improvements: 1)Parameters from up-stream single-modal pretraining are inadequately applied,without proper jointly fine-tuning in a down-stream multi-modal task. 2) Toselect descriptive parts of images, a simple late fusion is widely applied,regardless of priori knowledge that language-related information should beencoded into a common linguistic embedding space by stronger encoders. 3) Dueto diversity across products, their attribute sets tend to vary greatly, butcurrent approaches predict with an unnecessary maximal range and lead to morepotential false positives. To address these issues, we propose in this paper anovel approach to boost multi-modal e-commerce attribute value extraction viaunified learning scheme and dynamic range minimization: 1) Firstly, a unifiedscheme is designed to jointly train a multi-modal task with pretrainedsingle-modal parameters. 2) Secondly, a text-guided information rangeminimization method is proposed to adaptively encode descriptive parts of eachmodality into an identical space with a powerful pretrained linguistic model.3) Moreover, a prototype-guided attribute range minimization method is proposedto first determine the proper attribute set of the current product, and thenselect prototypes to guide the prediction of the chosen attributes. Experimentson the popular multi-modal e-commerce benchmarks show that our approachachieves superior performance over the other state-of-the-art techniques.

Boosting Multi-Modal E-commerce Attribute Value Extraction via Unified Learning Scheme and Dynamic Range Minimization