DRILL: Dynamic Representations for Imbalanced Lifelong Learning

TL;DR

DRILL is a novel lifelong learning architecture for NLP that uses self-organizing neural networks to mitigate forgetting in imbalanced, non-stationary data streams without prior task boundary knowledge.

Contribution

It introduces a biologically inspired self-organizing neural architecture for continual learning in NLP, outperforming existing methods in challenging data scenarios.

Findings

DRILL outperforms current methods in imbalanced, non-stationary data.

It effectively mitigates forgetting without prior task boundary knowledge.

First to apply self-organizing neural architecture in open-domain lifelong NLP learning.

Abstract

Continual or lifelong learning has been a long-standing challenge in machine learning to date, especially in natural language processing (NLP). Although state-of-the-art language models such as BERT have ushered in a new era in this field due to their outstanding performance in multitask learning scenarios, they suffer from forgetting when being exposed to a continuous stream of data with shifting data distributions. In this paper, we introduce DRILL, a novel continual learning architecture for open-domain text classification. DRILL leverages a biologically inspired self-organizing neural architecture to selectively gate latent language representations from BERT in a task-incremental manner. We demonstrate in our experiments that DRILL outperforms current methods in a realistic scenario of imbalanced, non-stationary data without prior knowledge about task boundaries. To the best of our knowledge, DRILL is the first of its kind to use a self-organizing neural architecture for open-domain lifelong learning in NLP.