NLU-STR at SemEval-2024 Task 1: Generative-based Augmentation and Encoder-based Scoring for Semantic Textual Relatedness

TL;DR

This paper presents a BERT-based approach for semantic textual relatedness in multiple Arabic dialects, achieving top rankings in SemEval-2024 shared tasks through generative augmentation and encoder-based scoring.

Contribution

It introduces a novel combination of generative augmentation and encoder-based scoring methods for multilingual semantic relatedness tasks.

Findings

Achieved 1st place in MSA with a Spearman score of 0.49

Ranked 5th for Moroccan dialect with a score of 0.83

Ranked 12th for Algerian dialect with a score of 0.53

Abstract

Semantic textual relatedness is a broader concept of semantic similarity. It measures the extent to which two chunks of text convey similar meaning or topics, or share related concepts or contexts. This notion of relatedness can be applied in various applications, such as document clustering and summarizing. SemRel-2024, a shared task in SemEval-2024, aims at reducing the gap in the semantic relatedness task by providing datasets for fourteen languages and dialects including Arabic. This paper reports on our participation in Track A (Algerian and Moroccan dialects) and Track B (Modern Standard Arabic). A BERT-based model is augmented and fine-tuned for regression scoring in supervised track (A), while BERT-based cosine similarity is employed for unsupervised track (B). Our system ranked 1st in SemRel-2024 for MSA with a Spearman correlation score of 0.49. We ranked 5th for Moroccan and 12th for Algerian with scores of 0.83 and 0.53, respectively.