Improving the editing process of automatically produced lecture transcripts based on Natural Language Analysis

Automatically produced lecture transcripts can act as an alternative to traditional note taking, benefiting those students whose needs and preferences are not met in the traditional learning environment. Nonetheless, despite the substantial progress that has been made in the area of Automatic Speech Recognition (ASR), the performance of ASR systems is still below the levels required for accurate transcription of lectures. This paper describes the development of a tool, which facilitates the evaluation of automatically produced transcription files, based on Natural Language Analysis. This tool is a step forward in the production of meaningful materials for disabled students, with minimal investment in time and effort by academic staff, thereby improving the accessibility of traditional teaching methodologies.