Visualization of Large Databases

Never before in history data has been generated at such high volumes as it is today. Exploring and analyzing the vast volumes of data becomes increasingly difficult. Information visualization and visual data mining can help to deal with the flood of information. The advantage of visual data exploration is that the user is directly involved in the data mining process. There is a large number of information visualization techniques which have been developed over the last decade to support the exploration of large data sets.

• Customer Analysis of Call Volume
  
  (In collaboration with AT&T)
  The visual exploration of massive data sets arising from telecommunication networks and services is a challenge. SWIFT-3D is an integrated data visualization and exploration system created at AT&T Labs for large scale network analysis.
  
  
  
• Data Analysis in Molecular Biology
  
  In recent years the amount of biological data that was collected has been constantly growing. However, the data is complex and therefore difficult to understand. Visual Exploration techniques can help to get valuable insights and to make the existing information accessible. In our project we focus on tools to explore aligned protein sequences and to support the discovery of correlated residues in proteins.
  
  
• Exploration of Electronic Mail
  
  Handling email has become an important task in almost every business process as well as in personal life. With the steady growth of email volumes, handling becomes a difficult task. Methods from the field of artificial intelligence try to automatically filter out the unwanted mail to reduce the overload. We presented a visual approach to cope with emails. The concept is not to replace automatic methods, but rather to enhance and supervise them. Our MailExplorer offers the user the capability to look at his mails in a geographic, a timely and a content-based manner.
  
  
• Exploration of Network Data Traffic
  
  Network communication has become indispensable in business, education, and government. With the pervasive role of the Internet as a means of sharing information across networks, its misuse for destructive purposes has grown immensely in the recent years. We work on interactive visualization techniques for gaining deeper insight into the network flow behavior by means of user-driven visual exploration. Integrating data warehouse technology, information visualization, and decision support, brings about the benefit of efficiently collecting the input data and aggregating over very large data sets, visualizing the results, and providing interactivity to facilitate analytical reasoning.
  
  
• Exploring and Visualizing Digital Libraries
  
  Finding and Extracting interesting and potentially useful information from large information spaces, like Digital Libraries, is a challenging task. In this project, we try to have a radial hierarchy visualization technique for analysing and exploring co-authorship in large digital document libraries. It provides co-authorship between authors in an intuitive way and the development of these relations over time in a single view.
  
  
• Pixel Bar Charts Project
  
  The basic idea is to use the pixels within the bars to present the detailed information of the data records. Our pixel bar charts retain the intuitiveness of traditional bar charts while allowing very large data sets to be visualized in an effective way.
  
  
  
• Route Map Visualization
  
  The subject of this project is the development of an algorithm for the representation of drive routes for way descriptions. Due to the fact that traditional route planners only provide scaled maps of a route, leading to important details being represented mostly too small and unimportant details being stretched on the whole map, this new algorithm should deliver optimal graphical way descriptions.
  
  
• VisDB System
  
  The VisDB has been developed to support the exploration of large databases. The VisDB system implements several visual data mining techniques, allowing an exploration of large databases (up-to about one million data values).
  
  
• Visual Data Mining for Information Fusion support
  
  The objective of this project consists of supporting the information fusion using a visual data mining system, that makes possible the interactive exploration of the existing data. The system is in charge of partitioning the data in homogenous object groups (cluster analyse).
  

  

  
  
• Visual Optimization
  
  The creation of timetables is an important practical problem. No conflicts are allowed if, for example, two lectures with the same professor cannot take place at the same time, and the wishes of the affected people should be taken into account as far as possible. Iterative time scheduling enables the visualization of the timetable and its relevant characteristics.
  

  

  
  
• Visualization of Data Cubes
  
  Many companies store data in so called Data Warehouses. These data are saved forming data cubes. The objective of this project consists of the creation of a visualization component for the data cubes of Lufthansa Systems Ltd.
  
  
• Visualization of E-Commerce Customer Data Sets
  
  Selling throuth Internet enables a quick and rational processing of orders. Working out the huge amount of data for the purchase process in the corresponding way and providing this information to the customers allow these customers to get an overview of the ordering process. Within the scope of our cooperation with the Hewlett Packard Research Labs in Palo Alto, California, new algorithms are beeing developed for the visual data mining of E-Commerce data sets.
  
  
• Visualization of High Dimensional Data
  
  Many applications need the clustering of large amounts of high dimensional data. The most automated clustering algorithms process high dimensional data in an uneffective way, that is, clusters with specific (unexpected) characteristics are not found. In order to solve this problem, new visual data mining techniques should be used. The basic idea consists of supporting the decisive steps of a further developed automated clustering process with visualization techniques.
  
  
• Visualization of Internet Access Data Sets
  
  Many companies use Internet to sell their products. This kind of internet presence is connected with the expenditure of considerable resources. In order to amortize these investments, large amounts of data are collected about the visitors of internet offers and their interactions. The aim of this project is to find an understandable visualization technique for the global visitor behaviour, the so-called clickstream, which allows an easy interpretation for the possible consequences for an internet offer.
  
  
• VisualPoints System
  
  In a large number of applications, data are collected and referenced by their spatial locations. Visualizing large amounts of spatially referenced data on a limited-size screen display often results in poor visualizations due to the high degree of overplotting of neighboring data points. The VisualPoints system implements a new approach to visualizing large amounts of spatially referenced data.