This paper describes our system designed for SemEval-2022 Task 8:
Multilingual News Article Similarity. We proposed a linguistics-inspired model
trained with a few task-specific strategies. The main techniques of our system
are: 1) data augmentation, 2) multi-label loss, 3) adapted R-Drop, 4) samples
reconstruction with the head-tail combination. We also present a brief analysis
of some negative methods like two-tower architecture. Our system ranked 1st on
the leaderboard while achieving a Pearson's Correlation Coefficient of 0.818 on
the official evaluation set.

本文介绍了我们的系统，以应对针对于 SemEval-2022 任务 8：多语言新闻文章相似性。我们提出了一种受语言学启发的模型，采用了几个任务特定的策略进行训练。我们的系统的主要技术包括：1）数据增强，2）多标签损失，3）适应 R-Drop，4）头尾组合中的样本重构。我们还对一些负面方法进行了简要分析，例如两塔架构。我们的系统在排行榜上排名第 1，并在官方评估集上实现了 0.818 的皮尔逊相关系数。