/verb_sel_code

Primary LanguageJupyter Notebook

Probabilistic Verb Selection for Data-to-Text Generation

This is the Python code for our experiments of the paper Probabilistic Verb Selection for Data-to-Text Generation, along with the datasets of verb-percentage pairs extracted from three corpora.

Prerequisites

Python 3.6+ and later, Anaconda 3.5 is better (the author used).

Introduction

AutomaticEvaluation: the experiment of automatic evaluation(MRR).

Data: the verb-percentage pairs extracted from three corpora.

Model: our proposed model and two baseline models.

Plot: the code (Jupyter Notebook) for the figure in the paper.

Questions: the code for generating questions of human evaluation.

Citation

In case of using the code, please cite the following paper:

Zhang D, Yuan J, Wang X, et al. Probabilistic verb selection for data-to-text generation[J]. Transactions of the Association for Computational Linguistics, 2018, 6: 511-527.

@article{zhang2018probabilistic,
  title={Probabilistic verb selection for data-to-text generation},
  author={Zhang, Dell and Yuan, Jiahao and Wang, Xiaoling and Foster, Adam},
  journal={Transactions of the Association for Computational Linguistics},
  volume={6},
  pages={511--527},
  year={2018},
}