Maximizing SLU Performance with Minimal Training Data Using Hybrid RNN Plus Rule-based Approach

Takeshi Homma, Adriano S. Arantes, Maria Teresa Gonzalez Diaz, Masahito Togami
2018 Proceedings of the 19th Annual SIGdial Meeting on Discourse and Dialogue  
Spoken language understanding (SLU) by using recurrent neural networks (RNN) achieves good performances for large training data sets, but collecting large training datasets is a challenge, especially for new voice applications. Therefore, the purpose of this study is to maximize SLU performances, especially for small training data sets. To this aim, we propose a novel CRF-based dialog act selector which chooses suitable dialog acts from outputs of RNN SLU and rule-based SLU. We evaluate the
more » ... ctor by using DSTC2 corpus when RNN SLU is trained by less than 1,000 training sentences. The evaluation demonstrates the selector achieves Micro F1 better than both RNN and rule-based SLUs. In addition, it shows the selector achieves better Macro F1 than RNN SLU and the same Macro F1 as rule-based SLU. Thus, we confirmed our method offers advantages in SLU performances for small training data sets.
doi:10.18653/v1/w18-5043 dblp:conf/sigdial/HommaADT18 fatcat:fwxcag23onh37bvj4tpexfreii