LX-CEFR is an online service for the quantitative analysis of texts along a range of linguistic metrics, and for the estimation of the proficiency level of texts.
These quantitative metrics are meant to provide suport in the classification of texts according to the proficiency levels in Quadro de Referência para o Ensino Português no Estrangeiro (QuaREPE). These metrics are not meant to be used to replace the decisions they are aimed at supporting.
This service is based on automatic text processing tools. Hence, the results it returns may not be always totally correct. Its high accuracy rate allows tough it to provide a useful service given its aim.
LX-CEFR is described on the following publication:
Branco, António, João Rodrigues, João Silva, Francisco Costa and Rui Vaz, 2014, "Assessing Automatic Text Classification for Interactive Language Learning" , Proceedings, iSociety2014 - IEEE International Conference on Information Society IEE, pp.72-80.