Deep Contextualized Self-training for Low Resource Dependency Parsing

By Guy Rotman, Roi Reichart. 2019

56 citations recorded by the source corpus.

Abstract Neural dependency parsing has proven very effective, achieving state-of-the-art results on numerous domains and languages. Unfortunately, it requires large amounts of labeled data, which is costly and laborious to create. In this paper we propose a self-training algorithm that alleviates this annotation bottleneck by training a parser on its own output. Our Deep Contextualized Self-training (DCST) algorithm utilizes representation models trained on sequence labeling tasks that are derived from the parser’s output when applied to unlabeled data, and integrates these models with the base parser through a gating mechanism. We conduct experiments across multiple languages, both in low resource in-domain and in cross-domain setups, and demonstrate that DCST substantially outperforms traditional self-training as well as recent semi-supervised training methods.1

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Natural Language Processing, Information Extraction

How researchers see this paper

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

In Artificial Intelligence

1. Alan Ramponi
2. Roi Reichart
3. Nadav Oved
4. Amir Feder
5. Barbara Plank
6. Uri Shalit
7. Chaehwa Yoo
8. Je-Won Kang

In Computer Vision and Pattern Recognition

1. Alan Ramponi
2. Amir Feder
3. Nadav Oved
4. Roi Reichart
5. Barbara Plank
6. Chaehwa Yoo
7. Je-Won Kang
8. Xiaofeng Liu

In Natural Language Processing

1. Barbara Plank
2. Alan Ramponi
3. Nadav Oved
4. Amir Feder
5. Roi Reichart
6. Uri Shalit
7. Yonatan Belinkov
8. Sebastian Ruder

In Information Extraction

1. Roi Reichart
2. Barbara Plank

Citations in the corpus

• CausaLM: Causal Model Explanation Through Counterfactual Language Models (2020, background): “n choosing pre-training tasks and understanding their effect (Jernite, Bowman, and Sontag 2017; Logeswaran and Lee 2018; Ziser and Reichart 2018; Dong et al. 2019; Chang et al. 2019; Sun et al. 2019; Rotman and Reichart 2019). The goal of this process is to generate a representation that captures va”
• Neural Unsupervised Domain Adaptation in NLP—A Survey (2020, background): “Pseudo-labeling has recently been studied for parsing with contextualized word representations (Rotman and Reichart, 2019; Lim et al., 2020) and a recent work proposes adaptive ensembling (Desai et al.”
• Deep Unsupervised Domain Adaptation: A Review of Recent Advances and Perspectives (2022, methodology): “The English OntoNotes 5.0 and the Universal Dependencies datasets are used for dependency parsing UDA evaluation (Rotman and Reichart, 2019).”