Computer Science

Graduate Research Forum Details

Personalized pedestrian navigation: user profile assessment under PC-RE framework

Abstract

Personalization has been applied and researched in the field of adaptive user interface, e-commerce, and requirements engineering. Many personalized systems don't work well because of difficulty in inferring users' characteristics, particularly in the early stage of application usage. I argue that the alternative is to pay attention to users' abilities, goals, and preferences at the very early stage of requirements engineering. The personal and contextual requirements (PC-RE) engineering, a three-layer framework with the focus on personal requirements analysis, accommodates individual characteristics and personal goals, which provides a good platform for personalization. I take a clinical approach towards personalization, and will focus on the user characteristics and requirements layer of PC-RE and investigate personal assessment (the first step towards personalization). I propose to use personalized pedestrian navigation system to demonstrate how individual profile assessment could be done under the PC-RE framework, because individuals exhibit considerable individual differences in spatial abilities and strategies related to navigation. I hypothesize that personalized navigation aids catered towards individual spatial abilities and strategies could provide better help in users' navigation. I discuss testing the hypothesis using a prototype in a "Wizard of Oz" way using the Cogbag system developed in our Wearable Computer Lab at the University of Oregon. In addition, I will discuss selection of personal spatial ability testing instrument and possible way to evaluate the studies.