Towards String-To-Tree Neural Machine Translation

By Roee Aharoni, Yoav Goldberg. 2017

157 citations recorded by the source corpus.

We present a simple method to incorporate syntactic information about the target language in a neural machine translation system by translating into linearized, lexicalized constituency trees. An experiment on the WMT16 German-English news translation task resulted in an improved BLEU score when compared to a syntax-agnostic NMT baseline trained on the same dataset. An analysis of the translations from the syntax-aware system shows that it performs more reordering during translation in comparison to the baseline. A small-scale human evaluation also showed an advantage to the syntax-aware system.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Graph Convolutional Encoders for Syntax-aware Neural Machine Translation (2017, background): “Aharoni and Goldberg (2017) propose neural string-to-tree by predicting linearized parse trees.”
Deep RNNs Encode Soft Hierarchical Syntax (2018, background): “However, previous work has shown that syntactic information (in the form of either input features or supervision) is useful for a wide variety of NLP tasks (Punyakanok et al., 2005; Chiang et al., 2009), even in the neural setting (Aharoni and Goldberg, 2017; Chen et al., 2017).”
Tree-to-tree Neural Networks for Program Translation (2018, background): “Other work study using neural networks to generate parse trees from input-output examples [11, 31, 1, 26, 32, 3, 12, 8, 7].”
Investigating Failures of Automatic Translation in the Case of Unambiguous Gender (2021, background): “For example, Eriguchi et al. (2016) model source-side syntax while Aharoni and Goldberg (2017) proposed models that generate linearized dependency trees.”
Search Engine Guided Non-Parametric Neural Machine Translation (2017, background): “…et al., 2015; Dong et al., 2015), multi-source translation (Zoph and Knight, 2016; Firat et al., 2016b; Garmash and Monz, 2016), multimodal translation (Caglayan et al., 2016) and linguistic knowledge guided translation (Nadejde et al., 2017; Aharoni and Goldberg, 2017; Eriguchi et al., 2017).”
Understanding and Improving Morphological Learning in the Neural Machine Translation Decoder (2017, methodology): “Aharoni and Goldberg (2017) proposed a method to predict the target sentence along with its syntactic tree.”
Neural Edit Completion (2020, unknown): “Path2Tree follows Aharoni and Goldberg [2017].”
Unsupervised Latent Tree Induction with Deep Inside-Outside Recursive Auto-Encoders (2019, methodology): “, 2018), machine translation (Aharoni and Goldberg, 2017; Eriguchi et al., 2017; Zaremoodi and Haffari, 2018), and text classification (Li and Roth, 2006; Tai et al.”
Novel positional encodings to enable tree-based transformers (2019, methodology): “Sequence-to-sequence models without explicit tree modeling have been applied to tree generation using only a simple linearization of the tree structure (Vinyals et al., 2015; Eriguchi et al., 2017; Aharoni & Goldberg, 2017).”
On internal language representations in deep learning: an analysis of machine translation and speech recognition (2018, background): “Syntactic trees may be added in different ways on the source side, in tree-to-string neural machine translation [63, 104], or on the target side, in string-to-tree translation [4, 314, 342].”
Improved Neural Machine Translation with Chinese Phonologic Features (2018, background): “For the target side, [1] proposed to replace the target sentence with the linearized, lexicalized constituency tree.”
Non-Monotonic Sequential Text Generation (2019, methodology): “In investigating this question, we are solely interested in considering non-monotonic generation that does not rely on external supervision, such as parse trees (Eriguchi et al., 2017; Aharoni & Goldberg, 2017).”
A Survey of Knowledge-enhanced Text Generation (2020, methodology): “title) [75, 149] Summarization Document [105, 146] [80, 84] [2, 104] [166] [65, 183] [72, 87] [84] [18, 47, 145] (e.g., an article) [43, 149] [81] Machine Sentence [152, 174] - [31, 124] - [103, 177] [1, 6] [15, 165] translation [131] Content Document [83] - - [41, 67, 150] - - - manipulation Conten”