Multi-Task Active Learning for Neural Semantic Role Labeling on Low Resource Conversational Corpus

TL;DR

This paper introduces a multi-task active learning framework combining Semantic Role Labeling and Entity Recognition to reduce data requirements and improve performance on Indonesian conversational data.

Contribution

It presents a novel multi-task active learning approach for SRL and ER, demonstrating efficiency gains and introducing a new Indonesian conversational SRL dataset.

Findings

Multi-task active learning outperforms single-task and standard multi-task learning.

Active learning reduces training data needs by 12%.

New Indonesian conversational SRL dataset introduced.

Abstract

Most Semantic Role Labeling (SRL) approaches are supervised methods which require a significant amount of annotated corpus, and the annotation requires linguistic expertise. In this paper, we propose a Multi-Task Active Learning framework for Semantic Role Labeling with Entity Recognition (ER) as the auxiliary task to alleviate the need for extensive data and use additional information from ER to help SRL. We evaluate our approach on Indonesian conversational dataset. Our experiments show that multi-task active learning can outperform single-task active learning method and standard multi-task learning. According to our results, active learning is more efficient by using 12% less of training data compared to passive learning in both single-task and multi-task setting. We also introduce a new dataset for SRL in Indonesian conversational domain to encourage further research in this area.