Information retrieval systems have become a part of almost all of our daily activities. And although these systems are becoming more and more user-friendly, a rapid increase in availability of information urges us to find new ways of searching through all this available data. Visualization of information is turning out to be one of the more promising solutions, as it offers a time effective and intuitive means of search. This master%s thesis aims to introduce information visualization through selected retrieval systems by determining which data is visually represented, which functions are supported by visualization and which visualization techniques are being used in the process. Thus, we tried to determine whether and to what extent information visualization is useful, effective and intuitive and what is its added value. Retrieval systems were evaluated through an expert study, and we used the design of criteria, scenarios, evaluation scale and comparison tables for a uniform and balanced comparison. Our results show considerable differences between retrieval systems in terms of data types, visualization techniques and functions supported by visualization. The frequency of visualization elements also varies among systems. Most of them use visualization only for displaying results, while some additionally include it in their search and browsing process. Yet even the more advanced systems are not immune to a certain amount of conservatism, especially when it comes to visualizing elements in searching, browsing and interactivity. Our findings suggest most improvements will be required exactly in afore mentioned fields, while special consideration should also be given to user-friendliness in terms of help, effectiveness and ease of use with different skill levels of users in mind. But above all, the quality of organisation and interconnectedness of data remains the first and crucial prerequisite for any information visualization.
|
