Papers by Michał Satława

1 papers
Multilingual Entity and Relation Extraction Dataset and Model (2021.eacl-main)

Copied to clipboard

Challenge: HERBERTa is a pipeline for a multilingual task involving two separate BERT models.
Approach: They propose a dataset and a model that combines two independently pretrained BERT models for a multilingual setting to approach the task of Joint Entity and Relation Extraction.
Outcome: The proposed dataset achieves micro F1 81.49 for English on the SMiLER dataset . the proposed pipeline is close to the current SOTA on CoNLL, SpERT .

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations