Visualization and XML Presentation
View the Powerpoint Slides
What is Visualization
One definition for visualization is a means to create a graphical representation of abstract information for human users. Using visualization techniques allow users to see information with a different view. Often, it can be easier to see trends and relationships in data by taking a step back and looking at the more general picture and to not get lost in the forest of individual data points. There are many new products in the social networking space that exemplify this point. Often companies or information-seekers gain new leads or possible sources upon seeing a clearly laid-out depiction of the people involved on a project and their associations with others. Graphical representations of large amounts of information help users more quickly understand “the big picture” by providing a more advanced level of presentation. Visualization can also provide a means for users to interact with information. Filters and dynamic graphics let the user manipulate the data presentation to more clearly show what the user wants to see. Recent product development across vendors shows increased capabilities for users to manipulate data sets with sliders as well as tools to help the user refine or conflate the original query on the client side using XML tools such as Ajax. Providing the ability to highlight, rearrange locations, and filter results lets users tailor result sets to more closely match their needs: “Fundamental factors for a good visualization interface are an overview of the structure for a global understanding; the ability to zoom and to select some nodes; and dynamic requests in order to filter data in real time.” (Le Grand, B., Soto, M. 2001)
Using Visualization
Information visualization can be used in different ways to create a better understanding of information. Viewing large and complex data files graphically can more clearly present relationships between different pieces of data, particularly with hierarchically structured files. For example, isolating one descendant by filtering out other descendants of a node in an XML file makes it much easier to trace the ancestry of the chosen descendant node. The root or source of the file structure is immediately apparent, and the level of depth in the structure is evident upon viewing the physical space between a selected node and the root.
Visualization can also be used to provide a brief graphical summary of a large data file. Mdplot, medical database plot, allows users to see a one page summary of database records using graphs that summarize the completeness of the database records and allow users to more quickly see and address problems.
Visualization tools also provide a means for a user to interact with search results. Tools such as Grokker dynamically change the presented results as the user refines a search, allowing searchers to quickly narrow down a search and find what they are looking for. Other tools graphically show how different pieces of information are related.
XTM and Topic Maps
Topic maps show the connections between different items or concepts allowing the user to see related terms and find new directions for a search. There are many forms of topic maps, but all of the different types attempt to demonstrate associations between named entities. XTM (XML Topic Maps) is the XML-based structure and syntax used to accomplish that goal. XTM is developed by a consortium and recently became an ISO standard.
There are three main tags in an XTM document: topics, associations, and occurrences. A topic is essentially the subject, or the name of the concept that is being described. It is the digital equivalent of some real-world idea or thing. It does not have to be a noun, and it can stand alone with no children elements or further description. An occurrence is a piece of information that is relevant to the topic. It could be a link, a string of text, or anything that expresses significance within that topic. One common example is that a link to an electronic version of Hamlet is an occurrence of Hamlet, the topic. An association is the relationship between two topics. Any combination of topics that have associations can be a topic map. Visual representations can use the relationships defined within an XTM document to display the networked whole of a data set.
Benefits of XML for Visualization
XML is well suited for use with visualization tools because of how it presents data to the visualization tools. One of the major benefits of XML is the separation of content from presentation, allowing visualization software to use the XML data without needing to account for other information that it does not need. Visualization software can show relationships between items by using XML tags to identify and categorize different pieces of information. Additionally, XML can present data with a hierarchical structure, letting visualization software further define these relationships.
Visualizing XML
Some visualization tools create graphical representations of XML files. These programs let users interact with the XML file and help them see through the verbosity of larger, more hierarchical files. Users can search for tags or text within nodes to quickly locate desired information. These tools provide methods to control what is shown by letting users position and rotate the file tree and expand, collapse and, filter nodes. XML visualization software give users a means to more clearly see file structures and allow users to trace node ancestry, see node relationships and modify XML files.
Examples: Hydra3d - http://hydra3d.sourceforge.net/ LukeLab Noder - http://lukelab.com/lab/noder/
Visualization Tools that Use XML
Aduna Cluster Map - http://aduna-software.com/products/technology/clustermap/index.html
Aduna Cluster Map uses XML data files to describe searchable information. The XML files contain information about the collected items as well as information that describe the content of the items and how to group them by subject. This tool presents searchable categories, letting the user choose which categories to look in during a search. The software then provides a cluster map of the results, grouping items by category and showing any overlap between multiple categories. This map allows users to see how different categories relate and can more easily locate specific items that belong to a particular combination of categories.
Aduna Cluster map supports dynamic searching, letting users add and remove categories as they see fit. The cluster map changes instantly without any addition al input required from the user. This visualization also lets users see “near misses” during searches. If a search for items that belong to a combination of three categories were to produce no valid results, the cluster map would show what two category combinations existed. The searchers could then alter the search parameters as needed and immediately see how those changes affected the results and more quickly find the needed information.
ThinkMap uses XML as one valid format for data sources. The developers also provide XML-based configuration files in their API, making it easy for outsiders to extend and quickly customize their instance of the application.
Kartoo - http://www.kartoo.comKartoo receives data from a server via an XML flow, much like RSS. If the data is not already structured, the engine can execute the query on the server side and build the result set of links, then send it in an XML flow back to the Flash interface of Kartoo.
Grokker - http://www.grokker.comGrokker relies on XML feeds from content sources. The application submits queries to different sources and receives XML feeds of results. Grokker then maps those results to OpenSearch, processes the results by normalizing the data and eliminating duplicates, then outputs OpenSearch XML to the visualization piece of the application to display.
Resources
Aduna Cluster Map Library version 2005.1: Integration Guide (2005). Retrieved April 29, 2006 from http://aduna-software.com/index.html
Brodlie, J. W. (2002) XML for Visualization, EuroWeb 2002 Conference. Oxford, UK.
Dugas, M. (2001). XML-based visualization of design and completeness in medical databases. Medical Informatics and the Internet in Medicine. 26(4), 237-250.
Geroimenko, V., Chen, C. (2002) Visualizing the Semantic Web : XML-based internet and information visualization Springer-Verlag, London.
Grokker + FAST : Maximize the value of your FAST investment. (2006). Retrieved April 28, 2006 from http://www.grokker.com
Le Grand, B., Soto, M. (2000) Information management – Topic Maps visualization, XML Europe 2000, Paris, France.
Mazza, Riccardo. (2004). Introduction to Information Visualisation. Retrieved May 7, 2006 from University of Lugano website: http://www.istituti.usilu.net/mazzar/Web/infovis_introduction.pdf
ThinkMap SDK v. 2.6 Technical Whitepaper. (2005). Retrieved May 1, 2006 from http://www.thinkmap.com/whitepapers.jsp;jsessionid=A64A55D0AE07A804E85F190F626A87D2.thinkmap1
