Paper describes the system developed by the Laboratoire d’analysestatistique des textes (LAST) for the Lexical Complexity Prediction shared taskat SemEval-2021 . The proposed system is made up of a LightGBM model fed withfeatures obtained from many word frequency lists, published lexical norms andpsychometric data . For tackling the specificity of the multi-word task, it usesbigram association measures .

Author(s) : Yves Bestgen

Links : PDF - Abstract

Code :

Keywords : word - complexity - semeval - system - measures -

Leave a Reply

Your email address will not be published. Required fields are marked *