UXmood - A Tool to Investigate the User Experience (UX) Based on Multimodal Sentiment Analysis and Information Visualization (InfoVis)

A Test Management System to Support Remote Usability Assessment of Web Applications

“…The variety and the complexity of the experience users could have on the web can be categorised in three sections, user's perception, i.e., the process of identifying and interpreting the stimuli of the external world through one's sensory organs, recognition, i.e., the process of determining an action plan to achieve a goal by evaluating the input information, and behaviour, i.e., the process of executing the action plan [12]. They also find that the use of different observation methods is effective in evaluating web usability, and that each experience demands a specific evaluation method, and in their totality for a comprehensive evaluation methodology which integrates self-reported (e.g., Questionnaire for User Interaction Satisfaction (QUIS) [27], Software Usability Measurement Inventory (SUMI) [8], Web Site Analysis and MeasureMent Inventory (WAMMI) [28] specifically developed for web platforms, and System Usability Scale (SUS) [9]), observational, and physiological evaluation methods [13][14][15]. Observational methods consist of non-intrusive means of acquiring behavioural data regarding the performance, the efficiency, and the errors while the users interact with the product.…”

“…It helps resolve issues regarding selfreported methods when the users are unaware of the observation. Observational methods include video-based facial expression analysis [11], emotion recognition from human voice [13], and tracking user interaction by logging user actions [14]. Facial expression analysis can be used to understand a person's emotional response by employing an emotion and valence model, where valence is the intrinsic attractiveness (positive) or averseness (negative) of a situation and is associated with the emotion being expressed [29].…”

“…Several solutions have been proposed in literature to ease the analysis process. For example, Franco et al [14] present UXmood, a tool that condenses audio, video, text, and eye-tracking data in a "dashboard of coordinated visualisations". This tool also collects logs of user interaction and allows for replaying sessions of tests while supporting brushing and details-on-demand interactions which are synchronised with a temporal slider, allowing moderators to see specific moments of the tests freely.…”

“…The authors of [12] classify the experience a user can have on a web platform based on the user's perception, cognition, and behaviour. They and other researchers [13][14][15] recall that user experience is measured by different constructs related to usability, user perception and human emotional response, and that each method by itself is not able to uncover the true emotional experience. The research conducted by [16] shows that various factors influence the use of one web platform over another, such as navigation, content, accessibility, or appearance.…”

“…Therefore, traditional tests are particularly limiting and problematic to organise, while the need to perform more and more remote tests (both synchronous and asynchronous) has been felt more than ever in the last two years, due to the spread of the COVID pandemic forcing many researchers to work remotely from their homes and imposing restrictions on visiting users. Current AI-based techniques and technologies enable usability practitioners to collect and analyse large amounts of data locally during system testing, such as audios and videos, and to record user emotions, shifting the focus to a different set of challenges, namely optimising convolutional networks and classification algorithms to achieve real-time performance [17], or collecting data from different devices, or improving data interpretation through multimodal channels [13,14]. These days, usability testing can be conducted with a laptop and the software that people use every day, allowing facilitators to work remotely.…”