Multi-Task Learning for Argumentation Mining in Low-Resource Settings
This addresses the challenge of limited data in argumentation mining for NLP researchers, though it appears incremental in testing MTL assumptions.
The study tackled the problem of argument component identification in low-resource settings by investigating multi-task learning (MTL), finding that MTL outperforms single-task learning when training data is scarce, with performance improvements noted in such scenarios.
We investigate whether and where multi-task learning (MTL) can improve performance on NLP problems related to argumentation mining (AM), in particular argument component identification. Our results show that MTL performs particularly well (and better than single-task learning) when little training data is available for the main task, a common scenario in AM. Our findings challenge previous assumptions that conceptualizations across AM datasets are divergent and that MTL is difficult for semantic or higher-level tasks.