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

Code-switching, also called code-mixing, is the linguistics phenomenon wherein casual settings, multilingual speakers mix words from different languages inone utterance. Due to its spontaneous nature, code-switching is extremelylow-resource, which makes it a challenging problem for language and speechprocessing tasks. In such contexts, Code-Switching Language Identification(CSLID) becomes a difficult but necessary task if we want to maximally leverageexisting monolingual tools for other tasks. In this work, we propose two novelapproaches toward improving language identification accuracy on anEnglish-Mandarin child-directed speech dataset. Our methods include a stackedResidual CNN+GRU model and a multitask pre-training approach to use AutomaticSpeech Recognition (ASR) as an auxiliary task for CSLID. Due to thelow-resource nature of code-switching, we also employ careful silver datacreation using monolingual corpora in both languages and up-sampling as dataaugmentation. We focus on English-Mandarin code-switched data, but our methodworks on any language pair. Our best model achieves a balanced accuracy of0.781 on a real English-Mandarin code-switching child-directed speech corpusand outperforms the previous baseline by 55.3%.

Simple yet Effective Code-Switching Language Identification with Multitask Pre-Training and Transfer Learning