English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

This paper aims to improve the performance of text-to-SQL parsing byexploring the intrinsic uncertainties in the neural network based approaches(called SUN). From the data uncertainty perspective, it is indisputable that asingle SQL can be learned from multiple semantically-equivalentquestions.Different from previous methods that are limited to one-to-onemapping, we propose a data uncertainty constraint to explore the underlyingcomplementary semantic information among multiple semantically-equivalentquestions (many-to-one) and learn the robust feature representations withreduced spurious associations. In this way, we can reduce the sensitivity ofthe learned representations and improve the robustness of the parser. From themodel uncertainty perspective, there is often structural information(dependence) among the weights of neural networks. To improve thegeneralizability and stability of neural text-to-SQL parsers, we propose amodel uncertainty constraint to refine the query representations by enforcingthe output representations of different perturbed encoding networks to beconsistent with each other. Extensive experiments on five benchmark datasetsdemonstrate that our method significantly outperforms strong competitors andachieves new state-of-the-art results. For reproducibility, we release our codeand data at https://github.com/AlibabaResearch/DAMO-ConvAI/tree/main/sunsql.

SUN: Exploring Intrinsic Uncertainties in Text-to-SQL Parsers