Abstract
The ambiguous annotation criteria bring into the divergence of Chinese Word Segmentation (CWS) datasets with various granularities. Multi-criteria learning leverage the annotation style of individual datasets and mine their common basic knowledge. In this paper, we proposed a domain adaptive segmenter to capture diverse criteria of datasets. Our model is based on Bidirectional Encoder Representations from Transformers (BERT), which is responsible for introducing external knowledge. We also optimize its computational efficiency via model pruning, quantization, and compiler optimization. Experiments show that our segmenter outperforms the previous results on 10 CWS datasets and is faster than the previous state-of-the-art Bi-LSTM-CRF model.
Abstract (translated)
歧义的注释标准导致了中文分词数据集在不同粒度上的差异。多准则学习利用单个数据集的注释样式，挖掘它们的共同基础知识。在本文中，我们提出了一个域自适应分割器来捕获不同的数据集标准。我们的模型是基于双向编码器表示从变压器（伯特），负责介绍外部知识。我们还通过模型修剪、量化和编译器优化来优化其计算效率。实验表明，我们的分段器在10个CWS数据集上优于以前的结果，并且比以前最先进的bi-lstm-crf模型更快。
URL
https://arxiv.org/abs/1903.04190
