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

The Bidirectional Encoder Representations from Transformers (BERT) model hasbeen radically improving the performance of many Natural Language Processing(NLP) tasks such as Text Classification and Named Entity Recognition (NER)applications. However, it is challenging to scale BERT for low-latency andhigh-throughput industrial use cases due to its enormous size. We successfullyoptimize a Query-Title Relevance (QTR) classifier for deployment via a compactmodel, which we name BERT Bidirectional Long Short-Term Memory (BertBiLSTM).The model is capable of inferring an input in at most 0.2ms on CPU. BertBiLSTMexceeds the off-the-shelf BERT model's performance in terms of accuracy andefficiency for the aforementioned real-world production task. We achieve thisresult in two phases. First, we create a pre-trained model, called eBERT, whichis the original BERT architecture trained with our unique item title corpus. Wethen fine-tune eBERT for the QTR task. Second, we train the BertBiLSTM model tomimic the eBERT model's performance through a process called KnowledgeDistillation (KD) and show the effect of data augmentation to achieve theresembling goal. Experimental results show that the proposed model outperformsother compact and production-ready models.

Deploying a BERT-based Query-Title Relevance Classifier in a Production System: a View from the Trenches