Abstracting User Metrics for Intelligent User Interfaces in MDD: A UsiXML Case Study
(2026) Computing — Vol. 109, n° 1 (2026)
Files
Computing_2025___Metricas (5).pdf
Open Access - Adobe PDF
- 955.73 KB
Details
- Authors
Gaspar, Alberto 
Panach, José Gil, Miriam
- Abstract
- Runtime adaptation of graphical user interface elements is a key requirement in the development of intelligent user interfaces. To enable such adaptation, systems must be able to identify user characteristics accurately and in a non-intrusive manner, allowing the construction of user profiles that support more comfortable, safe, and effective interaction. This paper proposes a set of user metrics for collecting interaction data and estimating user characteristics at runtime, grounded on a previously defined user model. Based on a targeted literature review, the metrics are classified into automatic and manual depending on whether their parameters can be captured during interaction. As a main contribution, we define the parameters required to operationalise each metric, enabling their systematic computation. We propose the definition of metrics for 19 UCs, four of which (knowledge, skills, device, and interests) were empirically validated through an experiment involving 60 users. The experiment aimed to evaluate the degree of alignment between the characteristics derived from the proposed metrics and the users' self-reported preferences. It included 15 interactive tasks across two scenarios involving two interactive applications: a home automation system and an autonomous car. The results show significant correlations for all the user characteristics studied, supporting the feasibility of using interaction-based metrics to inform adaptive user interface design. Overall, this work represents a step towards a methodological foundation for designing intelligent user interfaces that dynamically adjust to user behaviour and actions at runtime.
- Affiliations
Citations
Gaspar, A., Panach, J., Gil, M., & Vanderdonckt, J. (2026). Abstracting User Metrics for Intelligent User Interfaces in MDD: A UsiXML Case Study. Computing, 109(1). https://doi.org/10.1007/978-3-032-26836-5_35 (Original work published 2026)