Assessing Subjective Quality of Web Interaction with Neural Network as Context of Use Model

Abstract

Despite certain advances in automation of software quality assurance, testing and debugging remain the most laborious activities in the software development cycle. Evaluation of web interaction quality is still largely performed with traditional human effort-intensive methods, particularly due to the inevitable association of website usability with particular contexts of use, target users, tasks, etc. We believe that testing automation in this field may ultimately lead to better online experience for all and are important in promoting e-society development. We propose to employ artificial neural networks to predict website users’ subjective impressions, whose importance is widely recognized but that are somehow overshadowed by the effectiveness and efficiency dimensions. We justify the structure of the network, with the input layer reflecting context of use, while the output layer consisting of the subjective evaluation scales (Beautiful, Evident, Fun, Trustworthy, and Usable). The experimental session with 82 users and 21 university websites was undertaken to collect the evaluation data for the network training. Finally, we verify the validity of the model by comparing it to a certain baseline, analyze the importance of the input factors, and provide recommendations for future evaluations-collecting sessions.