MPEG-7 low level image descriptors for modeling users' web pages visual appeal opinion

摘要

The study of the users' web pages first impression is an important factor for interface designers, due to its influence over the final opinion about a site. In this regard, the analysis of web aesthetics can be considered as an interesting tool for evaluating this early impression, and the use of low level image descriptors for modeling it in an objective way represents an innovative research field. According to this, in this paper we present a new model for website aesthetics evaluation and users' opinion prediction based on the use of some parameters derived from the MPEG-7 standard, which are related to three different dimensions of the image: texture, chrominance and luminance. The obtained results confirms the power of our solution, as well as the usefulness of the model for reducing costs during the design and implementation process of new websites by providing a rapid feedback about the interface visual appeal.