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/

Zendy Plus

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

Zendy Tools

Zendy Open

Scoring a large number of candidates precisely in several milliseconds isvital for industrial pre-ranking systems. Existing pre-ranking systemsprimarily adopt the \textbf{two-tower} model since the ``user-item decouplingarchitecture'' paradigm is able to balance the \textit{efficiency} and\textit{effectiveness}. However, the cost of high efficiency is the neglect ofthe potential information interaction between user and item towers, hinderingthe prediction accuracy critically. In this paper, we show it is possible todesign a two-tower model that emphasizes both information interactions andinference efficiency. The proposed model, IntTower (short for\textit{Interaction enhanced Two-Tower}), consists of Light-SE, FE-Block andCIR modules. Specifically, lightweight Light-SE module is used to identify theimportance of different features and obtain refined feature representations ineach tower. FE-Block module performs fine-grained and early featureinteractions to capture the interactive signals between user and item towersexplicitly and CIR module leverages a contrastive interaction regularization tofurther enhance the interactions implicitly. Experimental results on threepublic datasets show that IntTower outperforms the SOTA pre-ranking modelssignificantly and even achieves comparable performance in comparison with theranking models. Moreover, we further verify the effectiveness of IntTower on alarge-scale advertisement pre-ranking system. The code of IntTower is publiclyavailable\footnote{https://github.com/archersama/IntTower}

IntTower: the Next Generation of Two-Tower Model for Pre-Ranking System