cs60075_team2 at SemEval-2021 Task 1 : Lexical Complexity Prediction using Transformer-based Language Models pre-trained on various text corpora

TL;DR

This paper explores fine-tuning transformer-based language models on various text corpora to improve lexical complexity prediction, achieving high correlation scores in SemEval 2021 tasks.

Contribution

It introduces domain-specific fine-tuning of transformers for lexical complexity prediction and analyzes the impact of different models and aggregation methods.

Findings

Achieved Pearson correlation of 0.784 in sub-task 1

Achieved Pearson correlation of 0.836 in sub-task 2

Demonstrated effectiveness of domain-specific corpora in complexity prediction

Abstract

This paper describes the performance of the team cs60075_team2 at SemEval 2021 Task 1 - Lexical Complexity Prediction. The main contribution of this paper is to fine-tune transformer-based language models pre-trained on several text corpora, some being general (E.g., Wikipedia, BooksCorpus), some being the corpora from which the CompLex Dataset was extracted, and others being from other specific domains such as Finance, Law, etc. We perform ablation studies on selecting the transformer models and how their individual complexity scores are aggregated to get the resulting complexity scores. Our method achieves a best Pearson Correlation of $0.784$ in sub-task 1 (single word) and $0.836$ in sub-task 2 (multiple word expressions).