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

Lexically constrained text generation aims to control the generated text byincorporating some pre-specified keywords into the output. Previous workinjects lexical constraints into the output by controlling the decoding processor refining the candidate output iteratively, which tends to generate genericor ungrammatical sentences, and has high computational complexity. To addressthese challenges, we propose Constrained BART (CBART) for lexically constrainedtext generation. CBART leverages the pre-trained model BART and transfers partof the generation burden from the decoder to the encoder by decomposing thistask into two sub-tasks, thereby improving the sentence quality. Concretely, weextend BART by adding a token-level classifier over the encoder, aiming atinstructing the decoder where to replace and insert. Guided by the encoder, thedecoder refines multiple tokens of the input in one step by inserting tokensbefore specific positions and re-predicting tokens with low confidence. Tofurther reduce the inference latency, the decoder predicts all tokens inparallel. Experiment results on One-Billion-Word and Yelp show that CBART cangenerate plausible text with high quality and diversity while significantlyaccelerating inference.

Parallel Refinements for Lexically Constrained Text Generation with BART