# Structured Training for Neural Network Transition-Based Parsing

arXiv: https://arxiv.org/abs/1506.06158 · HTML: https://mlsanity.com/paper/1506.06158/

## Key facts

- Cited by 11 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 19 Jun 2015.

## Built on by (most cited first, 11 of 11)

- [Identifying Morality Frames in Political Tweets using Relational Learning](https://mlsanity.com/paper/2109.04535/) (2021)
- [Learning Beam Search Policies via Imitation Learning](https://mlsanity.com/paper/1811.00512/) (2018)
- [Back-Training excels Self-Training at Unsupervised Domain Adaptation of Question Generation and Passage Retrieval](https://mlsanity.com/paper/2104.08801/) (2021)
- [Transformers as Graph-to-Graph Models](https://mlsanity.com/paper/2310.17936/) (2023)
- [Randomized Deep Structured Prediction for Discourse-Level Processing](https://mlsanity.com/paper/2101.10435/) (2021)
- [Generating CCG Categories](https://mlsanity.com/paper/2103.08139/) (2021)
- [Data Distribution-based Curriculum Learning](https://mlsanity.com/paper/2402.07352/) (2024)
- [Mapping the Course for Prompt-based Structured Prediction](https://mlsanity.com/paper/2508.15090/) (2025)
- [Revisiting Tri-training of Dependency Parsers](https://mlsanity.com/paper/2109.08122/) (2021)
- [Algorithms for Weighted Pushdown Automata](https://mlsanity.com/paper/2210.06884/) (2022)
- [A Semi-Autoregressive Graph Generative Model for Dependency Graph Parsing](https://mlsanity.com/paper/2306.12018/) (2023)

## Abstract (arXiv)

We present structured perceptron training for neural network transition-based dependency parsing. We learn the neural network representation using a gold corpus augmented by a large number of automatically parsed sentences. Given this fixed network representation, we learn a final layer using the structured perceptron with beam-search decoding. On the Penn Treebank, our parser reaches 94.26% unlabeled and 92.41% labeled attachment accuracy, which to our knowledge is the best accuracy on Stanford Dependencies to date. We also provide in-depth ablative analysis to determine which aspects of our model provide the largest gains in accuracy.