Multi-Reference Training with Pseudo-References for Neural Translation and Text Generation
arXiv:1808.09564
Abstract
Neural text generation, including neural machine translation, image captioning, and summarization, has been quite successful recently. However, during training time, typically only one reference is considered for each example, even though there are often multiple references available, e.g., 4 references in NIST MT evaluations, and 5 references in image captioning data. We first investigate several different ways of utilizing multiple human references during training. But more importantly, we then propose an algorithm to generate exponentially many pseudo-references by first compressing existing human references into lattices and then traversing them to generate new pseudo-references. These approaches lead to substantial improvements over strong baselines in both machine translation (+1.5 BLEU) and image captioning (+3.1 BLEU / +11.7 CIDEr).
arXiv comment10 pages
References in corpus (6)
- Neural Machine Translation by Jointly Learning to Align and Translate
- Sequence to Sequence Learning with Neural Networks
- Convolutional Sequence to Sequence Learning
- Sequence to Sequence -- Video to Text
- Classify or Select: Neural Architectures for Extractive Document Summarization
- Bootstrapping Lexical Choice via Multiple-Sequence Alignment