A Hierarchical Decoding Model For Spoken Language Understanding From Unaligned Data
This work addresses the problem of improving SLU systems for easier data annotation, though it is incremental as it builds on existing methods for unaligned data.
The paper tackles spoken language understanding from unaligned data by proposing a hierarchical decoding model that dynamically parses act-slot-value triples, outperforming previous state-of-the-art on the DSTC2 dataset with effective generalization to unseen pairs and out-of-vocabulary values.
Spoken language understanding (SLU) systems can be trained on two types of labelled data: aligned or unaligned. Unaligned data do not require word by word annotation and is easier to be obtained. In the paper, we focus on spoken language understanding from unaligned data whose annotation is a set of act-slot-value triples. Previous works usually focus on improve slot-value pair prediction and estimate dialogue act types separately, which ignores the hierarchical structure of the act-slot-value triples. Here, we propose a novel hierarchical decoding model which dynamically parses act, slot and value in a structured way and employs pointer network to handle out-of-vocabulary (OOV) values. Experiments on DSTC2 dataset, a benchmark unaligned dataset, show that the proposed model not only outperforms previous state-of-the-art model, but also can be generalized effectively and efficiently to unseen act-slot type pairs and OOV values.