MultiSpider: Towards Benchmarking Multilingual Text-to-SQL Semantic Parsing
Longxu Dou, Yan Gao, Mingyang Pan, Dingzirui Wang, Wanxiang Che, Dechen Zhan, Jian-Guang Lou
Abstract
Text-to-SQL semantic parsing is an important NLP task, which facilitates the interaction between users and the database. Much recent progress in text-to-SQL has been driven by large-scale datasets, but most of them are centered on English. In this work, we present MultiSpider, the largest multilingual text-to-SQL semantic parsing dataset which covers seven languages (English, German, French, Spanish, Japanese, Chinese, and Vietnamese). Upon MultiSpider we further identify the lexical and structural challenges of text-to-SQL (caused by specific language properties and dialect sayings) and their intensity across different languages. Experimental results under various settings (zero-shot, monolingual and multilingual) reveal a 6.1% absolute drop in accuracy in non-English languages. Qualitative and quantitative analyses are conducted to understand the reason for the performance drop of each language. Besides the dataset, we also propose a simple schema augmentation framework SAVe (Schema-Augmentation-with-Verification), which significantly boosts the overall performance by about 1.8% and closes the 29.5% performance gap across languages.
BibTeX
@article{Dou_Gao_Pan_Wang_Che_Zhan_Lou_2023, title={MultiSpider: Towards Benchmarking Multilingual Text-to-SQL Semantic Parsing}, volume={37}, url={https://ojs.aaai.org/index.php/AAAI/article/view/26499}, DOI={10.1609/aaai.v37i11.26499}, abstractNote={Text-to-SQL semantic parsing is an important NLP task, which facilitates the interaction between users and the database. Much recent progress in text-to-SQL has been driven by large-scale datasets, but most of them are centered on English. In this work, we present MultiSpider, the largest multilingual text-to-SQL semantic parsing dataset which covers seven languages (English, German, French, Spanish, Japanese, Chinese, and Vietnamese). Upon MultiSpider we further identify the lexical and structural challenges of text-to-SQL (caused by specific language properties and dialect sayings) and their intensity across different languages. Experimental results under various settings (zero-shot, monolingual and multilingual) reveal a 6.1% absolute drop in accuracy in non-English languages. Qualitative and quantitative analyses are conducted to understand the reason for the performance drop of each language. Besides the dataset, we also propose a simple schema augmentation framework SAVe (Schema-Augmentation-with-Verification), which significantly boosts the overall performance by about 1.8% and closes the 29.5% performance gap across languages.}, number={11}, journal={Proceedings of the AAAI Conference on Artificial Intelligence}, author={Dou, Longxu and Gao, Yan and Pan, Mingyang and Wang, Dingzirui and Che, Wanxiang and Zhan, Dechen and Lou, Jian-Guang}, year={2023}, month={Jun.}, pages={12745-12753} }