A Framework for Large-Scale Automatic Fluency Assessment

Abstract

Learning assessments are important to monitor the progress of students throughout the teaching process. In the digital era, many local and large-scale learning assessments are conducted through technological tools. In this view, a large-scale learning assessment can be designed to tackle one or multiple parts of the teaching process. Oral reading fluency assessments evaluate the ability to read reference texts. However, even though the use of applications to collect the reading of the students avoids logistics costs and speeds up the process, the evaluation of recordings has become a challenging task. Therefore, this work presents a computational solution for large-scale precision-critical fluency assessment. The goal is to build an approach based on automatic speech recognition (ASR) for the automatic evaluation of the oral reading fluency of children and reduce hiring costs as much as possible.