Improving academic writing with a method for text revision supported by text mining

Abstract

In spite of the undisputed importance of writing, especially in the learning context of higher education, we still find many students with difficul-ties to write, present arguments, and keep their writing coherent. Therefore, this research aims to investigate how a learning tool with text mining functions can support students in the process of academic writing, based on certain criteria for text analysis. A quasi-experimental study with control and experimental groups was carried out in order to evaluate the tool’s capacity to help students revise and improve their texts. The experimental group had access to the mining tool, which guided them in the revision of global features of their texts according to four principles: continuity, congruency, progression and non-contradiction. A group of lecturers corrected the students’ tasks using the same correction grid. Results showed that the students who used the mining tool achieved higher grades when compared to the students who did not use it. Providing one-to-one support for the revision of certain text facets at the global level has been one of the main contributions of this research.