RUM: An Approach to Support Web Applications Adaptation During User Browsing

摘要

In order to fulfill the needs and preferences of today's web users, adaptive Web applications have been proposed. Existing adap-tation approaches usually adapt the content of pages according to the user interest. However, the adaptation of the interface structure to meet user needs and preferences is still incipient. In addition, building adap-tive Web applications requires a lot of effort from developers. In this paper, we propose an approach to support the development of adap-tive Web applications, analyzing the user behavior during navigation, and exploring the mining of client logs. In our approach, called RUM (Real-time Usage Mining), user actions are collected in the application's interface and processed synchronously. Thus, we are able detect behavioral patterns for the current application user, while she is browsing the application. Facilitating its deployment, RUM provides a toolkit which allows the application to consume information about the user behavior. By using this toolkit, developers are able to code adaptations that are automatically triggered in response to the data provided by the toolkit. Experiments were conducted on different websites to demonstrate the efficiency of the approach to support interface adaptations that improve the user experience.