What question did this study set out to answer?

This research aims to develop BERT-based models for improving relation extraction and semantic role labeling.

April 10, 2019Open Access

Simple BERT Models for Relation Extraction and Semantic Role Labeling

Puntos clave

This research aims to develop BERT-based models for improving relation extraction and semantic role labeling.
Conducted extensive experiments using simple BERT models on relevant datasets.
Evaluated performance without external features, focusing solely on BERT capabilities.
Compared results against existing state-of-the-art models.
Achieved state-of-the-art performance in both relation extraction and semantic role labeling tasks.
Demonstrated that simple BERT models can outperform complex systems by effectively leveraging internal features.

Resumen

We present simple BERT-based models for relation extraction and semantic role labeling. In recent years, state-of-the-art performance has been achieved using neural models by incorporating lexical and syntactic features such as part-of-speech tags and dependency trees. In this paper, extensive experiments on datasets for these two tasks show that without using any external features, a simple BERT-based model can achieve state-of-the-art performance. To our knowledge, we are the first to successfully apply BERT in this manner. Our models provide strong baselines for future research.

Preguntar a la IA

Me gusta

Guardar

Ver artículo completo