IITK@LCP at SemEval 2021 Task 1: Classification for Lexical Complexity Regression Task

TL;DR

This paper presents a novel approach for lexical complexity prediction using ELECTRA and a combination of classification and regression models, incorporating weak supervision signals to improve accuracy.

Contribution

It introduces a hybrid modeling approach for lexical complexity regression tasks and leverages weak supervision signals from Gloss-BERT to enhance performance.

Findings

Achieved MAE of 0.0654 on Sub-Task 1

Achieved MAE of 0.0811 on Sub-Task 2

Weak supervision from Gloss-BERT significantly improved results

Abstract

This paper describes our contribution to SemEval 2021 Task 1: Lexical Complexity Prediction. In our approach, we leverage the ELECTRA model and attempt to mirror the data annotation scheme. Although the task is a regression task, we show that we can treat it as an aggregation of several classification and regression models. This somewhat counter-intuitive approach achieved an MAE score of 0.0654 for Sub-Task 1 and MAE of 0.0811 on Sub-Task 2. Additionally, we used the concept of weak supervision signals from Gloss-BERT in our work, and it significantly improved the MAE score in Sub-Task 1.