Détails Publication
Automatic Graphical User Interface aesthetic evaluation tool using the UIED segmentation algorithm,
Discipline: Informatique et sciences de l'information
Auteur(s): Yacouba Kyelem, Mamadou Zo, Kisito K. Kabore, Frédéric T. Ouedraogo
Auteur(s) tagués:
Renseignée par : KABORE Kiswendsida Kisito
Résumé

In this work, we deal with the issue of automating the evaluation of the aesthetics of the user's applications interfaces. The aesthetics of the user interface is a sub-factor of software quality. The aesthetics of the interface provides several advantages to the users, such as the attraction of the interface, an impact on the usability of the interface, a great pleasure of the interface's use, etc. That is why we deal with the evaluation of the aesthetics interfaces application. Our goal is to suggest a tool for quantitative evaluation of user interfaces using metrics. That is why we have developed a tool for evaluating the aesthetics of application interfaces that takes screenshots of mobile and desktop applications. This tool also generates metrics as output. For the segmentation of the different components of the interfaces we used the UIED algorithm which is an algorithm for the automatic segmentation of images. Our tool uses 10 of the 14 metrics of Ngo et al [1]. We generated the values with our tool by taking the images of a site. We used a total of 351 images. The validation of our tool was done by comparing the data calculated by our tool to those of QUESTIM. We used Pearson correlation test to perform the comparison and we obtained a correlation rate of 0.91. This means that our tool and QUESTIM evaluate the interfaces in almost the same way.

Mots-clés

Automatic Graphical User Interface, aesthetic evaluation , segmentation algorithm

935
Enseignants
7851
Publications
49
Laboratoires
101
Projets