PolyU-CBS at TSAR-2022 : a simple, rank-based method for complex word substitution in two steps

Abstract

In this paper, we describe the system that we presented at the TSAR-2022 shared task regarding Lexical Simplification for English, Portuguese, and Spanish. We proposed an unsupervised approach in two steps: First, we used a masked language model with word masking for each language to extract possible candidates for the replacement of the difficult word; in the second step, we ranked the candidates according to three different Transformer-based metrics. Finally, we determined our list of candidates based on the lowest average rank across different metrics.