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

Nature language inference (NLI) task is a predictive task of determining theinference relationship of a pair of natural language sentences. With theincreasing popularity of NLI, many state-of-the-art predictive models have beenproposed with impressive performances. However, several works have noticed thestatistical irregularities in the collected NLI data set that may result in anover-estimated performance of these models and proposed remedies. In thispaper, we further investigate the statistical irregularities, what we refer asconfounding factors, of the NLI data sets. With the belief that some NLI labelsshould preserve under swapping operations, we propose a simple yet effectiveway (swapping the two text fragments) of evaluating the NLI predictive modelsthat naturally mitigate the observed problems. Further, we continue to trainthe predictive models with our swapping manner and propose to use the deviationof the model's evaluation performances under different percentages of trainingtext fragments to be swapped to describe the robustness of a predictive model.Our evaluation metrics leads to some interesting understandings of recentpublished NLI methods. Finally, we also apply the swapping operation on NLImodels to see the effectiveness of this straightforward method in mitigatingthe confounding factor problems in training generic sentence embeddings forother NLP transfer tasks.

What If We Simply Swap the Two Text Fragments? A Straightforward yet  Effective Way to Test the Robustness of Methods to Confounding Signals in  Nature Language Inference Tasks