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 Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Recently, deep learning models have been widely spread in the industrialrecommender systems and boosted the recommendation quality. Though havingachieved remarkable success, the design of task-aware recommender systemsusually requires manual feature engineering and architecture engineering fromdomain experts. To relieve those human efforts, we explore the potential ofneural architecture search (NAS) and introduce AMEIR for Automatic behaviorModeling, interaction Exploration and multi-layer perceptron (MLP)Investigation in the Recommender system. The core contributions of AMEIR arethe three-stage search space and the tailored three-step searching pipeline.Specifically, AMEIR divides the complete recommendation models into threestages of behavior modeling, interaction exploration, MLP aggregation, andintroduces a novel search space containing three tailored subspaces that covermost of the existing methods and thus allow for searching better models. Tofind the ideal architecture efficiently and effectively, AMEIR realizes theone-shot random search in recommendation progressively on the three stages andassembles the search results as the final outcome. Further analysis revealsthat AMEIR's search space could cover most of the representative recommendationmodels, which demonstrates the universality of our design. The extensiveexperiments over various scenarios reveal that AMEIR outperforms competitivebaselines of elaborate manual design and leading algorithmic complex NASmethods with lower model complexity and comparable time cost, indicatingefficacy, efficiency and robustness of the proposed method.

AMEIR: Automatic Behavior Modeling, Interaction Exploration and MLP Investigation in the Recommender System