Multi-task Learning of Pairwise Sequence Classification Tasks Over Disparate Label Spaces
This addresses the challenge of leveraging unlabeled and auxiliary data for NLP tasks with disparate labels, though it is incremental as it builds on existing multi-task and semi-supervised methods.
The paper tackles the problem of jointly learning multiple sequence classification tasks with different label sets by combining multi-task and semi-supervised learning, resulting in state-of-the-art performance for topic-based sentiment analysis.
We combine multi-task learning and semi-supervised learning by inducing a joint embedding space between disparate label spaces and learning transfer functions between label embeddings, enabling us to jointly leverage unlabelled data and auxiliary, annotated datasets. We evaluate our approach on a variety of sequence classification tasks with disparate label spaces. We outperform strong single and multi-task baselines and achieve a new state-of-the-art for topic-based sentiment analysis.