Wei Er Luo
ABSTRACT
The goal of data-to-text generation is to generate fluent and fidelity natural language based on input data. At present, most models focus on data planning. Generating text according to the planning can effectively improve the fidelity to the generated text, but they lack attention to the fluency of the generated text, resulting in poor readability of the generated text. In order to address this issue, we propose a data-to-text generation model based on a pre-training language model. A control module is cleverly designed inside the language model to improve the fidelity to generated text. In training stage, multiple loss functions are cleverly designed to optimize the fluency and fidelity of generated text. Finally, the experimental results of RotoWire datasets show that the strong prior knowledge based on the Generative Pre-Training 2.0 (GPT2) can effectively improve the fluency of the generated text, and the proposed model in this paper can generate and control the fluency and fidelity of the generated text simultaneously.
- We received support for this work from Sichuan Provincial Key Research and Development Project No. 2023YFG0020, as well as Major Science and Technology Project of Sichuan Province No. 2022ZDZX0001.Google Scholar
- ReferencesGoogle Scholar
- [1] O. Vinyals, M. Fortunato and N. Jaitly. "Pointer Networks." Computer Science, vol. 28, no., pp., 2015.Google Scholar
- [2] R. Puduppully, L. Dong and M. Lapata. Data-to-Text Generation with Content Selection and Planning. 2018.Google Scholar
- [3] R. Puduppully and M. Lapata. "Data-to-text Generation with Macro Planning." vol., no., pp., 2021.Google Scholar
- [4] R. Puduppully, Y. Fu and M. Lapata. "Data-to-text Generation with Variational Sequential Planning." vol., no., pp., 2022.Google Scholar
- [5] K. Chen, F. Li, B. Hu , "Neural data-to-text generation with dynamic content planning." Knowledge-Based Systems, vol. 215, no., pp., doi:10.1016/j.knosys.2020.106610, 2021.Google ScholarCross Ref
- [6] X. Zhang, Y. Zou, H. Zhang Automatic product copywriting for e-commerce. In Proceedings of the Proceedings of the AAAI Conference on Artificial Intelligence, 2022; pp. 12423-12431.Google Scholar
- [7] Z. Chen, H. Eavani, Y. Liu and W.Y. Wang. "Few-shot NLG with Pre-trained Language Model." arXiv, vol., no., pp., 2019.Google Scholar
- [8] H. Gong, Y. Sun, X. Feng, B. Qin and T. Liu. TableGPT: Few-shot Table-to-Text Generation with Table Structure Reconstruction and Content Matching. In Proceedings of the Proceedings of the 28th International Conference on Computational Linguistics, 2020.Google ScholarCross Ref
- [9] A. Chan, Y.S. Ong, B. Pung, A. Zhang and J. Fu. CoCon: A Self-Supervised Approach for Controlled Text Generation. 2020.Google Scholar
- [10] L. Logeswaran and H. Lee. "An efficient framework for learning sentence representations." vol., no., pp., 2018.Google Scholar
- [11] S. Wiseman, S.M. Shieber and A.M. Rush. Challenges in Data-to-Document Generation. In Proceedings of the Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing, 2017.Google ScholarCross Ref
- [12] H. Gong, X. Feng, B. Qin and T. Liu. Table-to-Text Generation with Effective Hierarchical Encoder on Three Dimensions (Row, Column and Time). 2019.Google ScholarCross Ref
- [13] H. Gong, W. Bi, X. Feng, B. Qin and T. Liu. Enhancing Content Planning for Table-to-Text Generation with Data Understanding and Verification. In Proceedings of the Findings of the Association for Computational Linguistics: EMNLP 2020, 2020.Google ScholarCross Ref
Index Terms
- Data-to-text Generation with Data Control and Multi-loss Fusion
Recommendations
Neural Methods for Data-to-text Generation
The neural boom that has sparked natural language processing (NLP) research throughout the last decade has similarly led to significant innovations in data-to-text generation (D2T). This survey offers a consolidated view into the neural D2T paradigm with ...
Remodeling Numerical Representation for Text Generation on Small Corpus: A Syntactical Analysis
ACAI '19: Proceedings of the 2019 2nd International Conference on Algorithms, Computing and Artificial IntelligenceData-to-text generation aims to generate natural language descriptions from non-linguistic data. Recent research on data-to-text generation often uses a neural encoder-decoder architecture due to its simplicity to work across multiple domain. In this ...
Automatic Generation of Air Quality Index Textual Forecasts Using a Data-To-Text Approach
Proceedings of the 16th Conference of the Spanish Association for Artificial Intelligence on Advances in Artificial Intelligence - Volume 9422In this paper we present a data-to-text service which automatically produces textual forecasts about the air quality state for every municipality in Galicia NW Spain for the Galician Meteorology Agency MeteoGalicia. We discuss the context and the ...
Comments