Generating Natural Language Descriptions From Tables

Abstract

This paper proposes a neural generative architecture, namely NLDT, to generate a natural language short text describing a table which has formal structure and valuable information. Specifically, the architecture maps fields and values of a table to continuous vectors and then generates a natural language description by leveraging the semantics of a table. The NLDT architecture adopts a two-level neural model to make the most of the structure of a table to fully express the relationship between contents. To deal with the problem of out-of-vocabulary, this paper develops a simple and fast word-conversion method that replaces rare words appearing in texts with common field information in tables and directly replicates contents from table to the output sequence according to the field information. Besides, this paper adds the concept of theme to adapt the NLDT architecture to open domain and improves beam search algorithm to strengthen the results in the inference stage. On the WEATHERGOV dataset, the NLDT architecture improves the state-of-the-art BLEU-4 score from 61.01 to 62.89 and the current state-of-the-art F1 score from 73.21 to 78. On the WIKIBIO and WIKITABLE datasets, the NLDT architecture achieves a BLEU-4 score of 45.77 and 38.71 respectively which also outperform the state-of-the-art approaches. Furthermore, this paper introduces a Chinese dataset WIKIBIOCN including 33,244 biographies with corresponding tables. On the WIKIBIOCN dataset, the NLDT architecture achieves a BLEU-4 score of 38.87 and fairly good manual evaluation.

Language technologies impact on modern communication: analysis of new formats, challenges, and education

“…Wang, Hsiao, and Chang (2020) draw their attention to communicative language technology that assists in writing an automatic paper based on an RNN and the TextRank algorithm. Cao (2020) studies the process of generating natural language descriptions from different kinds of tables. Zhu et al (2020) outline modeling graph structure in a transformer for better abstract meaning representation to text generation.…”

Few-Shot Data-to-Text Generation via Unified Representation and Multi-Source Learning

“…More recent studies have utilized the capabilities of pre-trained language models in their designs, but have also incorporated specialized encoder structures or attention mechanisms specifically for table inputs. These include encoder-only models (Arik and Pfister, 2019;Yin et al, 2020;Herzig et al, 2020;Huang et al, 2020;Iida et al, 2021;Eisenschlos et al, 2021;Yang et al, 2022), as well as encoder-decoder models (Cao, 2020;Andrejczuk et al, 2022;. However, it should be noted that the encoder structures of these works are specifically tailored for table input and cannot be directly applied to other types of data.…”

A Brief Survey on Natural Language Processing Based Text Generation and Evaluation Techniques

“…In another research Cao et al [35] generates short text that explains a given table. A neural generative architecture known as NLDT was proposed by the researcher.…”