A Few-Shot Learning Focused Survey on Recent Named Entity Recognition and Relation Classification Methods

Abstract

Named Entity Recognition (NER) and Relation Classification (RC) are important steps in extracting information from unstructured text and formatting it into a machine-readable format. We present a survey of recent deep learning models that address named entity recognition and relation classification, with focus on few-shot learning performance. Our survey is helpful for researchers in knowing the recent techniques in text mining and extracting structured information from raw text.