Logic2Text: High-Fidelity Natural Language Generation from Logical Forms

Findings of the Association for Computational Linguistics: EMNLP 2020, pages 2096–2111 November 16 - 20, 2020. c⃝2020 Association for Computational Linguistics 2096 Logic2Text: High-Fidelity Natural Language Generation from Logical Forms Zhiyu Chen1, Wenhu Chen1, Hanwen Zha1, Xiyou Zhou1, Yunkai Zhang1, Sairam Sundaresan2, and William Yang Wang1 1University of California, Santa Barbara 2Intel AI {zhiyuchen, wenhuchen, hwzha, xiyou, yunkai zhang, william}@cs.ucsb.edu, sairam.sundaresan@intel.com Abstract Previous studies on Natural Language Gen- eration (NLG) from structured data have pri- marily focused on surface-level descriptions of record sequences. However, for complex structured data, e.g., multi-row tables, it is often desirable for an NLG system to de- scribe interesting facts from logical inferences across records. If only provided with the table, it is hard for existing models to pro- duce controllable and high-fidelity logical gen- erations. In this work, we formulate high- fidelity NLG as generation from logical forms in order to obtain controllable and faithful generations. We present a new large-scale dataset, LOGIC2TEXT, with 10,753 descrip- tions involving common logic types paired with the underlying logical forms. The logical forms show diversified graph struc- ture of free schema, which pose great chal- lenges on the model’s ability to understand the semantics. We experiment on (1) Fully- supervised training with the full datasets, and (2) Few-shot setting, provided with hun- dreds of paired examples; We compare sev- eral popular generation models and analyze their performances. We hope our dataset can encourage research towards building an ad- vanced NLG system capable of natural, faith- ful, and human-like generation. The dataset and code is available at https://github. com/czyssrs/Logic2Text. 1 Introduction Natural language generation (NLG) from struc- tured data has been an important research problem in many applications. Recent data-driven methods have achieved good performances on various NLG tasks (Liu et al., 2018; Freitag and Roy, 2018; Chen et al., 2019b). However most studies focus on sur- face descriptions of simple record sequences, for example, attribute-value pairs of fixed or very lim- ited schema, like E2E (Novikova et al., 2017) and WikiBio (Lebret et al., 2016). In real-world cases for multi-row tables, it is often more desirable and plausible to provide descriptions involving higher- level logical inference across data records. For example, in Figure 1, instead of plain restatements, human readers would be more favorable to abstract descriptions that can summarize or conclude infor- mation over the table records. To produce such logical-level generations of high fidelity, it is not yet appropriate to provide only the table as the input in a real-world NLG system, based on the following reasons: 1) Low Fidelity. Given only the table, it is challenging for existing neural models to produce such logically correct generations involving rea- soning and symbolic calculations, e.g., max, min, counting, averaging, etc. 2) Uncontrollable content selection. Given a table, the space of logically entailed descrip- tions is exponentially large, due to vast number of combinations of different operations and argu- ments from the table, e.g., count, comparison, superlative, etc. It is hard and uncontrol- lable for neural models to …

Content truncated to 500 words