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

User behavior modeling is important for industrial applications such asdemographic attribute prediction, content recommendation, and targetadvertising. Existing methods represent behavior log as a sequence of adopteditems and find sequential patterns; however, concrete location and timeinformation in the behavior log, reflecting dynamic and periodic patterns,joint with the spatial dimension, can be useful for modeling users andpredicting their characteristics. In this work, we propose a novel model basedon graph neural networks for learning user representations from spatiotemporalbehavior data. A behavior log comprises a sequence of sessions; and a sessionhas a location, start time, end time, and a sequence of adopted items. Ourmodel's architecture incorporates two networked structures. One is a tripartitenetwork of items, sessions, and locations. The other is a hierarchical calendarnetwork of hour, week, and weekday nodes. It first aggregates embeddings oflocation and items into session embeddings via the tripartite network, and thengenerates user embeddings from the session embeddings via the calendarstructure. The user embeddings preserve spatial patterns and temporal patternsof a variety of periodicity (e.g., hourly, weekly, and weekday patterns). Itadopts the attention mechanism to model complex interactions among the multiplepatterns in user behaviors. Experiments on real datasets (i.e., clicks on newsarticles in a mobile app) show our approach outperforms strong baselines forpredicting missing demographic attributes.

Calendar Graph Neural Networks for Modeling Time Structures in Spatiotemporal User Behaviors