Maximization of users experience in websites: estimating the optimum size of the multimedia content

In this paper, an extended experimental research is performed in order to calculate the mean opinions score (MOS) of several users considering quality of experience variables such as information perception and multimedia type of preference. The experimental campaign resulted in a great archive of measurements of the MOS of users over several multimedia files. It was found that multimedia type affects information perception whereas users prefer information to be presented by certain multimedia types. A quality factor is introduced for the first time for each multimedia type at the best of the authors' knowledge. This quality factor connects MOS (i.e. user quality experience) and file size (i.e. network resources) with multimedia type. These quality factors express a measure of how MOS is affected by each individual multimedia type and it is inherent to the multimedia type. Approximating equations incorporating MOS and file size as well as quality factors are extracted and are used to construct an optimization problem. The optimization problem has as a target to optimize (i.e. maximize) the overall MOS of information perception and multimedia type of preference of a website that present information using several multimedia types, constrained by the total size of the website in bytes. Finally, an upper bound for the MOS is extracted based on the total file size of the website in terms of bytes.