Lecture segmentation and transcription

School: Engineering

Team Members: James Hopgood, Ilie Galit, James Hanratty

Abstract

This student-led project, principally designed by two third year Electronic and Computer Science students, aims to enhance lecture recordings of live classes by developing software-based tools for semi-automatically segmenting lecture recordings into shorter videos and hosting them on a meta-enhanced multimedia channel alongside corresponding lecture notes.    The project is inspired by work in other Schools, such as videos for online courses in the School of Informatics. Creating such courses is often viewed as labour intensive but are one-off recordings. We intend to develop tools that enable lecturers to help automatically partition live lectures and teaching materials into a contemporary format.

The proposed benefits would be:

1. Semi-automatically segmented videos that students find easier to digest and focus on, giving structure to the online experience;

2. encouraging lecturers to organize their lectures in a more efficient topic-based manner;

3. organized content by topic with transcriptions (e.g. YouTube), allowing students to quickly find content;

4. the availability this resource gives academics the option of providing a flipped-classroom in subsequent years; the availably of these lecture segments also provides useful resources to lecturers who are covering temporary staff absence.

Overall, we aim to address the question of whether this content presentation will improve the student experience of lecture recording. We will engage key colleagues to use the tools developed and, based on feedback from staff and student, appropriately extend the scope of the project.