论文笔记丨Entity-Relation Extraction as Multi-turn Question Answering

---

作者：凯
单位：燕山大学

---

Abstract

先前实体关系抽取都是基于如下数据格式

[“cowden”, “'s”, “syndrome”, “is”, “an”, “autosomal”, “dominant”, “multiorgan”, “hamartoma”, “syndrome”, “characterized”, “by”, “benign”, “and”, “malignant”, “thyroid”, “tumors”, “,”, “breast”, “cancers”, “,”, “and”, “colon”, “cancers”, “.”], “h”: [“cowden’s syndrome”, “C0018553”, [[0, 1, 2]]], “t”: [“autosomal dominant”, “C0443147”, [