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

Sentence representation from vanilla BERT models does not work well onsentence similarity tasks. Sentence-BERT models specifically trained on STS orNLI datasets are shown to provide state-of-the-art performance. However,building these models for low-resource languages is not straightforward due tothe lack of these specialized datasets. This work focuses on two low-resourceIndian languages, Hindi and Marathi. We train sentence-BERT models for theselanguages using synthetic NLI and STS datasets prepared using machinetranslation. We show that the strategy of NLI pre-training followed by STSbfine-tuning is effective in generating high-performance sentence-similaritymodels for Hindi and Marathi. The vanilla BERT models trained using this simplestrategy outperform the multilingual LaBSE trained using a complex trainingstrategy. These models are evaluated on downstream text classification andsimilarity tasks. We evaluate these models on real text classification datasetsto show embeddings obtained from synthetic data training are generalizable toreal datasets as well and thus represent an effective training strategy forlow-resource languages. We also provide a comparative analysis of sentenceembeddings from fast text models, multilingual BERT models (mBERT, IndicBERT,xlm-RoBERTa, MuRIL), multilingual sentence embedding models (LASER, LaBSE), andmonolingual BERT models based on L3Cube-MahaBERT and HindBERT. We releaseL3Cube-MahaSBERT and HindSBERT, the state-of-the-art sentence-BERT models forMarathi and Hindi respectively. Our work also serves as a guide to buildinglow-resource sentence embedding models.

L3Cube-MahaSBERT and HindSBERT: Sentence BERT Models and Benchmarking BERT Sentence Representations for Hindi and Marathi