Deep Semi-Supervised Learning with Linguistically Motivated Sequence Labeling Task Hierarchies
This addresses sequence labeling in NLP, offering incremental improvements for tasks like Chunking.
The paper tackles semi-supervised learning for sequence labeling tasks by exploiting a linguistically motivated hierarchy to regularize supervised tasks with unsupervised error backpropagation, resulting in up to two percentage points F1 improvement for Chunking compared to a baseline.
In this paper we present a novel Neural Network algorithm for conducting semi-supervised learning for sequence labeling tasks arranged in a linguistically motivated hierarchy. This relationship is exploited to regularise the representations of supervised tasks by backpropagating the error of the unsupervised task through the supervised tasks. We introduce a neural network where lower layers are supervised by junior downstream tasks and the final layer task is an auxiliary unsupervised task. The architecture shows improvements of up to two percentage points F1 for Chunking compared to a plausible baseline.