Just-in-Time Cognitive Assessment and Task Recommendation for Individuals with Cognitive Impairments
Sean-Ryan Smith · 2015 · Proceedings of the 12th International Web for All Conference (W4A) · doi:10.1145/2745555.2746670
Summary
This doctoral consortium paper proposes a mobile system that performs just-in-time cognitive assessment for people with cognitive impairments (CI), then uses those assessment results to recommend tasks the individual is most likely to be able to perform independently. The core idea is that cognitive ability fluctuates throughout the day and across contexts — a person may be more capable of certain mental tasks at some times than others. Rather than relying on periodic clinical assessments, the system would deliver brief, non-intrusive mini-assessments at context-specific times tied to the user's daily schedule (e.g., medication time, mealtimes, recreation). By mapping cognitive strengths to specific moments and activities, the system could recommend which tasks a user is most likely to complete successfully, supporting independence while avoiding frustration from attempting tasks beyond their current capacity. The research targets stroke survivors and individuals with CI recruited through Cabrillo College's Stroke and Disability Learning Center. Pilot work had already been conducted on web-based brain-training software for stroke survivors, producing an initial set of design guidelines tested with five participants using a low-fidelity prototype.
Key findings
The proposed system would collect data from smartphone sensors (accelerometer, GPS) and apps (calendar entries) to learn temporal patterns in the user's daily routine and predict optimal times for assessment delivery. A data mining algorithm would combine these inputs to infer user activities and contexts. The system follows a participatory design methodology with requirement gathering through individual interviews, focus groups, and home visits with 10 participants to understand cognitive challenges, desired independent tasks, preferred delivery interfaces (phones, smartwatches), and contexts of use (home, workplace, rehabilitation centre). The biggest technical challenge identified is determining the right combination and timing of cognitive assessments — since task recommendations are only as valid as the assessment results they are based on. Prior pilot work with stroke survivors found that brain-training software needed specific design accommodations: direct personal contact with facilitators was important, and guidelines were needed for making rehabilitation software both effective and usable for this population. The anticipated contributions include enhanced guidelines for designing intuitive web and mobile systems for people with CI and solutions for mobile app accessibility problems specific to this population.
Relevance
This proposal addresses a meaningful gap in cognitive accessibility: the disconnect between static, clinic-based cognitive assessments and the dynamic, context-dependent nature of cognitive ability in daily life. For practitioners, the just-in-time assessment concept highlights that cognitive impairment is not a fixed state — people with CI may be capable of different tasks at different times depending on fatigue, medication, stress, and environment. This has implications for how we design adaptive interfaces and cognitive support tools. The participatory design approach, working directly with stroke survivors and people with CI through a disability learning centre, models good practice for inclusive research. However, this is a dissertation proposal rather than a completed study — no system has been built or evaluated, and the sample size of 10 participants is small. The data mining approach (combining accelerometer, GPS, and calendar data) raises privacy considerations that the paper does not address. The concept of context-aware, adaptive cognitive support remains an active and promising research area.
Tags: cognitive impairment · cognitive assessment · cognitive rehabilitation · assistive technology · mobile accessibility · stroke · independent living · wearable technology · participatory design