As mobile phones become a necessity to the daily lives of most people, there is a need to find an effective solution to the issues highlighted previously. As the population is also graying in several countries, there may be more users who will encounter difficulties with the interfaces of contemporary mobile phones.

Currently, there is no reported mobile phone engine that is able to learn and automatically reconfigure the interface widgets based on a user’s context. As an example, a user’s context could be defined by time, location, traffic condition, phone usage pattern and even personal biometrics like heart/perspiration rate etc. Currently, widgets will reside on the mobile phone interface even if the user has not used the applications for an extended period of time. These widgets will remain until the user specifically removes the application. Similarly, when new applications are installed, the user can either manually organize the corresponding widgets or just allow the widget “clutter” to get worse. Reasons for the lack of such capability could include anticipated high processing overheads, development constraints and potential increased access complexity introduced by such engines.

With an efficient context-aware reconfiguration engine, however, users need not expend considerable effort to navigate and locate a particular application as the system will automatically select, retrieve and display the icons most likely to be used for that user. An intelligent reconfiguration engine will also learn from the perceived context and suitably adapt the interface for the user doing away with frequent manual reconfiguration. Challenges facing the development of such an engine would then include:

• Need for a simple and efficient design that would not introduce unnecessary overheads or more complexity to the user interface.
• The engine must be transparent to the user while working in tandem with the operating system of the mobile phone
• The engine must learn and automatically adapt the interface according to the specified context

This paper proposes an intelligent, context-aware interface reconfiguration engine prototype for the mobile phone called SmartlFace and is organized as follows. Section 1 provides the background of the existing situation and associated problems while Section 2 presents the motivation and challenges behind the research. Section 3 presents a review of related research. Design details of the system architecture and learning engine are in Sections 4 and 5. Section 6 presents the system test results as well as a performance analysis and evaluation. The paper concludes with Section 7 followed by the references.