@inproceedings{song-etal-2020-structural,
    title = "Structural Information Preserving for Graph-to-Text Generation",
    author = "Song, Linfeng  and
      Wang, Ante  and
      Su, Jinsong  and
      Zhang, Yue  and
      Xu, Kun  and
      Ge, Yubin  and
      Yu, Dong",
    editor = "Jurafsky, Dan  and
      Chai, Joyce  and
      Schluter, Natalie  and
      Tetreault, Joel",
    booktitle = "Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics",
    month = jul,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/Author-page-Marten-During-lu/2020.acl-main.712/",
    doi = "10.18653/v1/2020.acl-main.712",
    pages = "7987--7998",
    abstract = "The task of graph-to-text generation aims at producing sentences that preserve the meaning of input graphs. As a crucial defect, the current state-of-the-art models may mess up or even drop the core structural information of input graphs when generating outputs. We propose to tackle this problem by leveraging richer training signals that can guide our model for preserving input information. In particular, we introduce two types of autoencoding losses, each individually focusing on different aspects (a.k.a. views) of input graphs. The losses are then back-propagated to better calibrate our model via multi-task training. Experiments on two benchmarks for graph-to-text generation show the effectiveness of our approach over a state-of-the-art baseline."
}