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

Domain shift is a big challenge in NLP, thus, many approaches resort tolearning domain-invariant features to mitigate the inference phase domainshift. Such methods, however, fail to leverage the domain-specific nuancesrelevant to the task at hand. To avoid such drawbacks, domain counterfactualgeneration aims to transform a text from the source domain to a given targetdomain. However, due to the limited availability of data, such frequency-basedmethods often miss and lead to some valid and spurious domain-tokenassociations. Hence, we employ a three-step domain obfuscation approach thatinvolves frequency and attention norm-based masking, to mask domain-specificcues, and unmasking to regain the domain generic context. Our experimentsempirically show that the counterfactual samples sourced from our masked textlead to improved domain transfer on 10 out of 12 domain sentimentclassification settings, with an average of 2% accuracy improvement over thestate-of-the-art for unsupervised domain adaptation (UDA). Further, our modeloutperforms the state-of-the-art by achieving 1.4% average accuracy improvementin the adversarial domain adaptation (ADA) setting. Moreover, our model alsoshows its domain adaptation efficacy on a large multi-domain intentclassification dataset where it attains state-of-the-art results. We releasethe codes publicly at \url{https://github.com/declare-lab/remask}.

ReMask: A Robust Information-Masking Approach for Domain Counterfactual Generation