Pre-trained models have achieved success in Chinese Short Text Matching (STM) tasks, but they often rely on superficial clues, leading to a lack of robust predictions. To address this issue, it is crucial to analyze and mitigate the influence of superficial clues on STM models. Our study aims to investigate their over-reliance on the edit distance feature, commonly used to measure the semantic similarity of Chinese text pairs, which can be considered a superficial clue. To mitigate STM models' over-reliance on superficial clues, we propose a novel resampling training strategy called Gradually Learn Samples Containing Superficial Clue (GLS-CSC). Through comprehensive evaluations of In-Domain (I.D.), Robustness (Rob.), and Out-Of-Domain (O.O.D.) test sets, we demonstrate that GLS-CSC outperforms existing methods in terms of enhancing the robustness and generalization of Chinese STM models. Moreover, we conduct a detailed analysis of existing methods and reveal their commonality.
@article{arxiv.2309.04162,
title = {GLS-CSC: A Simple but Effective Strategy to Mitigate Chinese STM Models' Over-Reliance on Superficial Clue},
author = {Yanrui Du and Sendong Zhao and Yuhan Chen and Rai Bai and Jing Liu and Hua Wu and Haifeng Wang and Bing Qin},
journal= {arXiv preprint arXiv:2309.04162},
year = {2023}
}