MPEG-DASH parametrisation for adaptive online streaming of different MOOC videos categories

The Dynamic Adaptive Streaming over HTTP (MPEG-DASH) ensures online videos display of good quality and without interruption. It provides an adequate streaming for each display device and network transmission. This can be verfield of Massive Open Online Courses (MOOCs). In fact, MPEG-DASH tracks the bandwidth fluctuations so hat the learner profits from continuous streaming, without worrying about frequent interruption of courses videos. Therefore, the learners profit from an exceptional visual experience that improves their commitment level and eases the course assimilation. These MPEG-DASH assets can become more and more advantageous if a good choice of its parameters is made. Being a recent branch, the MPEG-DASH adaptive diffusion presents a research field where the efforts are still limited, even more for MOOCs videos. Most of the work published in this sense focus on the Quality of Service (QoS) and the technical specifications of the network transmission. In this paper, we aim to consider the quality of the streamed content that directly impacts the learners Quality of Experience (QoE). For this, we develop a content-aware dataset that includes several MOOCs videos of different characteristics and types. These videos are firstly encoded using the latest codecs then dashified according to a coding scheme of several combinations of bitrates and display resolutions. Then, and in order to enhance the learners QoE, the so generated MPEG-DASH manifest files and segments are subsequently exploited to study the most appropriate codecs, bitrates and segment durations for each type of MOOCs videos.