A Pattern Recognition Approach for the Analysis of Lexical Availability in Math Teaching Program Students

Abstract

The lexical availability studies allow the characterization and analysis of the available lexicon of a group of individuals. Current analysis approaches focus mainly on index descriptions that summarize group characteristics of the words, without offering an analysis focused on individuals. The objective of the present work is to propose a space-vector model for the representation and analysis of the lexicons obtained from lexical availability surveys, allowing to complement the lexical availability studies, through a model of lexical corpus-based representation based on the traditional information retrieval models. The feasibility of the proposed model has been demonstrated through the use of pattern recognition techniques such as vector support and clustering machines in a case study of lexical availability in teaching students from two Chilean universities. In this research it is concluded that it is possible to use these techniques for the visualization, grouping and classification of students according to their available lexicon.