Visual Querying and Exploration of Large Answers in XML Databases with : A Demonstration Holger Meuss Center for Information and Language Processing University of Munich, Germany meuss@cis.uni-muenchen.de Klaus U. Schulz Center for Information and Language Processing University of Munich, Germany Franc¸ois Bry Institute for Computer Science University of Munich, Germany Abstract This paper describes (eXploring XML data), a sys- tem for efficiently querying XML data and visually explor- ing answer sets, with a special focus on the novel user in- terface for interaction with the query result. The process of querying XML documents is performed interactively and in a completely graphical manner. Special attention was paid to simplicity of the user interface: In a first step the user submits a query to the system and gets the result pre- sented in the form of a Complete Answer Aggregate (CAA). CAAs visually (re)present the answer space in a compact form and thus provide a good overview over the set of all answers. They offer a rich variety of presentation modes and exploration techniques which are used in the second step to interactively explore the answer space. 1 System architecture and query language The overall architecture of is depicted in Figure 1. The main components of are a query engine and the CAA engine. The query engine evaluates a query submitted by the user and computes the resulting CAA. It uses ded- icated index structures that guarantee efficient evaluation. After query evaluation is completed, the resulting CAA can be explored by interaction with the CAA engine. Since XML data is tree-structured data, we use a simple language for querying trees that extends Kilpel¨ ainen’s Tree Matching model [3]. The user draws a tree-shaped query (Figure 2) that will be matched with the tree-structured XML data. The user has the possibility to specify direct and indirect containment of elements, attribute values, text con- tainment, element names, order and direct neighborhood of User Index and Database XML Query Engine CAA Engine Request Import Data CAA Query Manipulation CAA Request Data Figure 1. System architecture of elements, and optional elements. The query language is de- scribed in more detail in [6]. 2 Complete Answer Aggregates A Complete Answer Aggregate (CAA), as depicted in Figure 3, is a datastructure that stores for every query node all matching database elements, together with their contain- ment relations. Figure 3 shows an example CAA for a query searching for cities with their respective sightseeing infor- 1