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

Current voice assistants typically use the best hypothesis yielded by theirAutomatic Speech Recognition (ASR) module as input to their Natural LanguageUnderstanding (NLU) module, thereby losing helpful information that might bestored in lower-ranked ASR hypotheses. We explore the change in performance ofNLU associated tasks when utilizing five-best ASR hypotheses when compared tostatus quo for two language datasets, German and Portuguese. To harvestinformation from the ASR five-best, we leverage extractive summarization andjoint extractive-abstractive summarization models for Domain Classification(DC) experiments while using a sequence-to-sequence model with a pointergenerator network for Intent Classification (IC) and Named Entity Recognition(NER) multi-task experiments. For the DC full test set, we observe significantimprovements of up to 7.2% and 15.5% in micro-averaged F1 scores, for Germanand Portuguese, respectively. In cases where the best ASR hypothesis was not anexact match to the transcribed utterance (mismatched test set), we seeimprovements of up to 6.7% and 8.8% micro-averaged F1 scores, for German andPortuguese, respectively. For IC and NER multi-task experiments, whenevaluating on the mismatched test set, we see improvements across all domainsin German and in 17 out of 19 domains in Portuguese (improvements based onchange in SeMER scores). Our results suggest that the use of multiple ASRhypotheses, as opposed to one, can lead to significant performance improvementsin the DC task for these non-English datasets. In addition, it could lead tosignificant improvement in the performance of IC and NER tasks in cases wherethe ASR model makes mistakes.

Using multiple ASR hypotheses to boost i18n NLU performance