Skip to main content

Table 1 Comparison between different RE methods

From: Subsequence and distant supervision based active learning for relation extraction of Chinese medical texts

Method type

Characteristics

Advantages and Disadvantages

Pipeline

Separate subtasks

Propagation error and reductant entities prediction

Parameter sharing

Merge subtasks

Void propagation error but cause computing resources problem

Sequence tagging

Merge subtasks

Reduce resources waste but need high-quality data

Sequence tagging and active learning

Merge subtasks and active learning

High-quality data problem is alleviated