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

Learned embeddings for products are an important building block for web-scalee-commerce recommendation systems. At Pinterest, we build a single set ofproduct embeddings called ItemSage to provide relevant recommendations in allshopping use cases including user, image and search based recommendations. Thisapproach has led to significant improvements in engagement and conversionmetrics, while reducing both infrastructure and maintenance cost. While mostprior work focuses on building product embeddings from features coming from asingle modality, we introduce a transformer-based architecture capable ofaggregating information from both text and image modalities and show that itsignificantly outperforms single modality baselines. We also utilize multi-tasklearning to make ItemSage optimized for several engagement types, leading to acandidate generation system that is efficient for all of the engagementobjectives of the end-to-end recommendation system. Extensive offlineexperiments are conducted to illustrate the effectiveness of our approach andresults from online A/B experiments show substantial gains in key businessmetrics (up to +7% gross merchandise value/user and +11% click volume).

ItemSage: Learning Product Embeddings for Shopping Recommendations at Pinterest