Análisis de la riqueza léxica en el contexto de la clasificación de atributos demográficos latentes

Abstract

In this paper we analyse the utility of lexical richness measures for predicting latent user attributes from Spanish opinionated texts. Our aim is to know how useful could be lexical richness to predict user's gender, age and regional origin. To this end, we applied 32 lexical richness measures over 1911 previously labeled texts with demographic information. This approach has the advantage that it is domain-independent with modest computational cost.