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

News feed recommendation is an important web service. In recent years,pre-trained language models (PLMs) have been intensively applied to improve therecommendation quality. However, the utilization of these deep models islimited in many aspects, such as lack of explainability and being incompatiblewith the existing inverted index systems. Above all, the PLMs basedrecommenders are inefficient, as the encoding of user-side information willtake huge computation costs. Although the computation can be accelerated withefficient transformers or distilled PLMs, it is still not enough to make timelyrecommendations for the active users, who are associated with super long newsbrowsing histories.  In this work, we tackle the efficient news recommendation problem from adistinctive perspective. Instead of relying on the entire input (i.e., thecollection of news articles a user ever browsed), we argue that the user'sinterest can be fully captured merely with those representative keywords.Motivated by this, we propose GateFormer, where the input data is gated beforefeeding into transformers. The gating module is made personalized, lightweightand end-to-end learnable, such that it may perform accurate and efficientfiltering of informative user input. GateFormer achieves highly impressiveperformances in experiments, where it notably outperforms the existingacceleration approaches in both accuracy and efficiency. We also surprisinglyfind that even with over 10-fold compression of the original input, GateFormeris still able to maintain on-par performances with the SOTA methods.

GateFormer: Speeding Up News Feed Recommendation with Input Gated Transformers