Characterization of semantic complexity in TOEFL passages

Abstract

Reading comprehension is a process that involves the realization of mental models, as well as the storage of memories in long-term memory; if we add to this that many students need to accredit other languages, it involves performing more cognitive processes. With mental models, it is possible to analyze the internal structures present, for example, in the reading comprehension section of the TOEFL test where three types of texts are involved: narrative, expository and argumentative. In this work, the three types of texts are semantically modeled as structures, where semantic relations are the input for their construction, and the semantic complexity depending on text type to be reviewed. By having these structures, students can graphically locate the elements that characterize each type of text, and an orientation is given to generate expert systems for the training of the activities to be developed.