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

We propose ConGraT(Contrastive Graph-Text pretraining), a general,self-supervised method for jointly learning separate representations of textsand nodes in a parent (or ``supervening'') graph, where each text is associatedwith one of the nodes. Datasets fitting this paradigm are common, from socialmedia (users and posts), to citation networks over articles, to link graphsover web pages. We expand on prior work by providing a general,self-supervised, joint pretraining method, one which does not depend onparticular dataset structure or a specific task. Our method uses two separateencoders for graph nodes and texts, which are trained to align theirrepresentations within a common latent space. Training uses a batch-wisecontrastive learning objective inspired by prior work on joint text and imageencoding. As graphs are more structured objects than images, we also extend thetraining objective to incorporate information about node similarity andplausible next guesses in matching nodes and texts. Experiments on variousdatasets reveal that ConGraT outperforms strong baselines on various downstreamtasks, including node and text category classification and link prediction.Code and certain datasets are available athttps://github.com/wwbrannon/congrat.

ConGraT: Self-Supervised Contrastive Pretraining for Joint Graph and Text Embeddings