Design and Evaluation of Computing Technologies to Promote Learning of Feature-Rich Software

点击翻译按钮获取中文摘要

Feature-rich software applications, such as photo-editing software, word-processing applications, and spreadsheet applications, are used by over one billion people each year to accomplish diverse productivity-related and recreational tasks. These applications are very powerful, in part owing to the hundreds (or even thousands) of commands and options that they make available. At the same time, the sheer volume of commands can make feature-rich software difficult for individuals to learn and use. Consequently, techniques for promoting efficient access to needed commands and for increasing the ease with which users learn feature-rich software have received considerable attention by the HCI and applied AI communities over the past 30 years. While this prior research has contributed several important techniques and design guidelines, the landscape of feature-rich software continues to shift substantially, raising a host of new challenges and opportunities including: a decentralization of the help resources upon which users rely, the increasing availability of web-based knowledge on command-to-task mappings, the increasing popularity of user-contributed content, and the growing diversity of computer users which now also includes young children.****The research I propose focuses on three main themes. The first theme is to create and evaluate technologies designed to improve users' access to and use of web-based tutorials by leveraging contributions and opinions of other tutorial followers. The second theme is to examine the feasibility and utility of providing users with in-application, contextually relevant command recommendations based on command groupings mined from web-based tutorials and other web-documentation. The third theme is to explore the problem of software complexity from the perspective of young children, a rapidly growing but understudied population of technology users. ****The proposed research will be done by applying user-centred design and interface prototyping methods, targeted applications of machine learning and natural language processing techniques (in collaboration with domain experts), and mixed-methods evaluations with human subjects. The outcomes of my proposed research will be novel interface designs and prototypes that improve the experience of using feature-rich software together with empirically validated design guidelines that synthesize the lessons learned from evaluating our solutions with end users.**