Exploring User Experience with Image Schemas, Sentiments, and Semantics

Abstract

Although the concept of user experience includes two key aspects, experience of meaning (usability) and experience of emotion (affect), the empirical work that measures both the usability and affective aspects of user experience is currently limited. This is particularly important considering that affect could significantly influence a user's perception of usability. This paper uses image schemas to quantitatively and systematically evaluate both these aspects. It proposes a method for evaluating user experience that is based on using image schemas, sentiment analysis, and computational semantics. The aim is to link the sentiments expressed by users during their interactions with a product to the specific image schemas used in the designs. The method involves semantic and sentiment analysis of the verbal responses of the users to identify (i) task-related words linked to the task for which a certain image schema has been used and (ii) affect-related words associated with the image schema employed in the interaction. The main contribution is in linking image schemas with interaction and affect. The originality of the method is twofold. First, it uses a domain-specific ontology of image schemas specifically developed for the needs of this study. Second, it employs a novel ontology-based algorithm that extracts the image schemas employed by the user to complete a specific task and identifies and links the sentiments expressed by the user with the specific image schemas used in the task. The proposed method is evaluated using a case study involving 40 participants who completed a set task with two different products. The results show that the method successfully links the users' experiences to the specific image schemas employed to complete the task. This method facilitates significant improvements in product design practices and usability studies in particular, as it allows qualitative and quantitative evaluation of designs by identifying specific image schemas and product design features that have been positively or negatively received by the users. This allows user experience to be assessed in a systematic way, which leads to a better understanding of the value associated with particular design features.

Culture Beneath Discourse: A Conceptual Model for Analyzing Nondeclarative Cultural Knowledge

“…However, my model illustrates many levels of cultural knowledge simultaneously at work, and several studies show that priming can take place at all of these levels (e.g. Boroditsky and Ramscar 2002;Sato et al 2015;Setchi and Asikhia 2019). Williams and Bargh (2008) show how the embodied experience of physical warmth can promote "warm" interpersonal judgments.…”

UXmood—A Sentiment Analysis and Information Visualization Tool to Support the Evaluation of Usability and User Experience

“…Among the analyzed works, only a few [8,30,31] use multimodal sentiment analysis techniques to evaluate user experience or usability. The more common type of classification is textual, whether in audio transcription or product reviews on the Internet [1,8,[30][31][32][33][34]. References [8,30,31,33,35] record the user's face for use in sentiment analysis algorithms based on facial expressions.…”

“…References [8,30,31,33,35] record the user's face for use in sentiment analysis algorithms based on facial expressions. User speech capture is present in six works [8,[30][31][32][33]36] but only four works [8,30,31,36] use audio features to perform sentiment analysis.…”

“…Only two studies [8,36] use sensors, such as eye trackers, to identify the area that the user is observing. Almost half of the studies [1,2,8,32,33,[41][42][43][44][45][46] use some form of questionnaire to obtain qualitative information about user experience or usability. The vast majority of studies perform in loco evaluation, with the specialist observing the test in the same location as the participant, except for References [8,30,36], where the specialist remotely observe user interaction.…”

On-body Sensing and Signal Analysis for User Experience Recognition in Human-Machine Interaction

“…Experience of emotion is one of the key aspects of user experience affecting to all aspects of the human-machine interaction (HMI), including utility, ease of use, and efficiency [1], the machine's ability to recognize user's emotion during user-machine interaction would improve the overall HMI usability. The machines that are aware of the user's emotion could adapt their activity features such as speed based on user's emotional state.…”