Dynamically Composing Domain-Data Selection with Clean-Data Selection by “Co-Curricular Learning” for Neural Machine Translation
Year: 2019
Type: conference-paper
Abstract: Noise and domain are important aspects of data quality for neural machine translation.Existing research focus separately on domaindata selection, clean-data selection, or their static combination, leaving the dynamic interaction across them not explicitly examined.This paper introduces a "co-curricular learning" method to compose dynamic domain-data selection with dynamic clean-data selection, for transfer learning across both capabilities.We apply an EM-style optimization procedure to further refine the "co-curriculum".Experiment results and analysis with two domains demonstrate the effectiveness of the method and the properties of data scheduled by the cocurriculum.
DOI: 10.18653/v1/p19-1123
Authors: Wei Wang, Isaac Caswell, Ciprian Chelba
Institution: Google (United States)
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W2952474700
