A Framework for Improving Adaptive Data Visualization in Decision Support Systems

Adaptive approaches are used to improve user experience and satisfaction for web browsing, based on profiling information gathered from user interactions. In decision support systems, the need for personalization adaptation has increased in order to provide more immediate and relevant information to decision makers, using web based access to data. Using visualizations for rendering complex query results, in real-time is of particular importance in many application domains. In this thesis we propose an approach, and a framework, for measuring history, experiences and satisfaction of users of a healthcare decision support system. The focus is on user selections of visualizations, based on the nature of queries generated. The aim of this framework is intended to provide collection of individual user experiences and satisfaction, in order to obtain a user population profile for later studies. The model used is a weighting scheme, but is designed to support later extensions and enhancements using \u27AI reasoning techniques\u27. This model was implemented and a usability study was conducted to validate improvements compared to non adaptive data visualization systems. The outcome of this research may lead to increased accuracy and reduced time of selection of visualization, over repeated usage, and is therefore important as a productivity enhancement approach