• Keine Ergebnisse gefunden

INVISIP : Metadata-based Information Visualization Techniques to Access Geodata Archives and to Support the Site Planning Process

N/A
N/A
Protected

Academic year: 2022

Aktie "INVISIP : Metadata-based Information Visualization Techniques to Access Geodata Archives and to Support the Site Planning Process"

Copied!
4
0
0

Wird geladen.... (Jetzt Volltext ansehen)

Volltext

(1)

CODATA Workshop

Information Visualization, Presentation, and Design 10-11 July 2002

INVISIP: Metadata-based Information Visualization Techniques to Access Geodata Archives and to Support the Site Planning Process

Stefan Göbel1, Jörg Haist1, Harald Reiterer2and Frank Müller2

1Fraunhofer Institute for Computer Graphics, GIS department, Fraunhoferstr. 5, 64283 Darmstadt, Germany

{stefan.goebel, joerg.haist}@igd.fhg.de

2University of Konstanz, Department of Computer & Information Science, Universitätsstr. 10, 78457 Konstanz, Germany

{harald.reiterer, frank.mueller}@uni-konstanz.de Introduction

Complex application examples or scenarios such as site planning need a lot of information, especially spatially referenced data (studies have shown that the amount of spatial information/geodata is 80-85% relating to all new data produced by the actual information society) in order to find a best place for a new building or an industrial area and to solve individual planning tasks such as generating ecological, environmental or traffic reports.

Recent initiatives to geospatial data archives or any other necessary information for the site planning process offer access to a wealth of distributed data covered by the widespread information spectrum of different geodata disciplines (e.g. environmental data, geologic data, cadastral data, remote sensing data or socio-demographic data), but offer only basic levels of interactivity and user assistance. The EU funded project INVISIP (Information Visualization for Site Planning) addresses this lack of usability and aims to develop new visualizations to support all involved parties within the different phases of the multi-step site planning process.

This paper introduces the GeoCrystal and the SuperTable + Scatterplot as new information visualization techniques (based on metadata and metaphors) to support accessing geospatial data archives and site planning in general.

INVISIP Framework

INVISIP has been initiated as technical project within the EU IST program, key action “III.4.2 information visualization” which aims to facilitate information access, data handling and navigation in (sometimes unfamiliar) information spaces. The proposed technical architecture (see figure 1) is generic in the sense that it can be used in various application fields. However, the actual implementation within the project concentrates on site planning, in particular traffic, infrastructure and environment-related issues are given special attention. From the technical point of view, the most innovative aspect within the INVISIP approach focuses on the development and usage of information visualization [1, 2] concepts and mechanisms to improve information retrieval and data mining aspects. The information visualization components of the INVISIP framework are integrated into a GeoLibrary as 3D information and navigation space, which uses metaphors enabling users to navigate in information spaces using intuitive spatial representations such as maps or well-known symbols settled in the library world like the book metaphor, information rooms or the complete library as 3D information and navigation space. The following chapters describe the structure and contents of the GeoCrystal and SuperTable + Scatterplot as new information visualization techniques to improve usability of the metadata browser within the INVISIP context. Both presented tools First publ. as paper in: 3rd CODATA Euro- American Workshop on Information Visualization Presentation and

Design, July 2002

Konstanzer Online-Publikations-System (KOPS) URL: http://www.ub.uni-konstanz.de/kops/volltexte/2007/3187/

URN: http://nbn-resolving.de/urn:nbn:de:bsz:352-opus-31873

(2)

are integrated into the GeoLibrary in which all components communicate among each other.

Based on the infrastructure of the InGeo Information Center [6] and its theme-overlapping metadata information system for geodata [4] a personalized metadata browser (see figure 1) will be established providing search and visualization functionality enhanced by metadata [5]

and further 2D/3D information visualization techniques and metaphors such as maps, information landscapes or books and libraries [8, 10, 12]. The metadata browser will help planning engineers to search for appropriate data (e.g. socio-demographic and economic data, geospatial and environmental data) and offers interfaces to the different components of the analyzer as add-on software component to support planning engineers using local planning software.

Data Acquisition

Metadata

Creation Local

Planning Tools Metadata Browser

Search & Visualise

Analyser I Visual Data Mining Metadata Information System

Data Integration II Metadata Entry

Data Integration I Data Warehouse

Analyser III Instrument metadata orig. data

Analyser II Repository Data

Acquisition Metadata

Creation Local

Planning Tools Metadata Browser

Search & Visualise

Analyser I Visual Data Mining Metadata Information System

Data Integration II Metadata Entry

Data Integration I Data Warehouse

Analyser III Instrument metadata orig. data

Analyser II Repository

Figure 1. INVISIP framework Geocrystal

GeoCrystal [3] has been developed as new information visualization technique combining algebraic mechanisms (InfoCrystal [12]) with library (book, bookshelves, floors, etc.) and spatial metaphors (2D maps respectively 3D spatial information landscapes). The global aim of GeoCrystal is to improve understanding of search results by offering familiar metaphors settled in the field of set theory (algebra) combined with the provision of illustrated visual parameters such as spatial references of result sets. Figure 2 provides two snapshots of GeoCrystal embedded into a Java 3D library for geodata. Hence the structure of GeoCrystal consists of two parts:

• On the ground of the information room there is a vector-based map realized as GIS component, which both enables users to specify spatial queries and supports spatial orientation and navigation.

• Above the map the result sets of a query are visually presented by block symbols indicating GeoCrystal elements, whereby the result sets are classified by the number of matching parameters corresponding to the query. Result sets matching all query parameters are located in the center of the map at an idea about the numerous distributions of results referring to special query parameters and how to modify the query in order to get better results. Similar to InfoCrystal each query parameter is mapped to a color and represented by a circle at the corners of the GeoCrystal (on ground level closely placed to the interactive map). Result set classes are visually mapped to other symbols such as rectangular (matching two parameters), triangular (matching three parameters) or square blocks (matching four parameters) and are colored according to the colors of matching parameters.

Activating the GeoCrystal initiates highlighting of the GeoCrystal element (class) containing the dataset. In order to compare several matching search hits, user can select the checkboxes in first column of a result list and activate a “comparison button”. Thus, another visualization is dynamically created offering detailed ranks such as a spatial rank, a thematic rank or a

(3)

temporal rank to visualize the precision of individual search parameters (distance between a value of the geodata attribute/metadata element and the query condition of a search parameter). At this, GeoCrystal provides visualizations such as tile bars, scatterplots or proportional symbols.

Figure 2. GeoCrystal embedded into a 3D information room of a library for geodata Supertable + Scatterplot

The goal of SuperTable + ScatterPlot [13] is to improve user experience during the information seeking process. The new visualizations are based on experiences of the authors developing a visual information retrieval system called INSYDER [7,10] to supply small and medium size enterprises with business information from the Internet. Based on extensive user tests the original visualizations (ResultTable, BarGraph, and SegmentView) have been redesigned in two different design variants, whereby only one is shown in Figure 3. Instead of offering multiple visualizations to choose from, the SuperTable + Scatterplot combines them in a new way. Therefore, the user has the feeling that he is working with one single visualization in different states. The second (Figure 3, left) and the third level of detail (Figure 3, right) are presented below. Further the SuperTable solves a problem, which seemed to be immanent to visualizations in document retrieval: the change of modalities.

Figure 3. Two peculiarities of the SuperTable + ScatterPlot.

In the second design variant, this problem is further smoothened by introducing the concept of granularity. It will be realized as six different levels of detail ranging from a simple 3-column view containing the relevance, an abstract text representation and the slider (to change the granularity) to a full text representation with the keywords highlighted. The SuperTable + ScatterPlot will be integrated into the library for geodata which will be used as 3D information

(4)

and navigation environment of the INVISIP metadata browser.

Summary & Outlook

This paper presents the GeoCrystal and SuperTable + ScatterPlot , which aims to support users in the data mining process of complex application areas such as site planning and to facilitate graphic-interactive access to geodata archives. These techniques are introduced as new information visualization techniques to support users during the different information retrieval phases, especially search result presentation and comparison of search results (hints for query modification).

From the application-oriented point of view, future work will be investigated to extend the GeoLibrary by application-driven access variants. Another main research focus will be to integrate further information visualization techniques within the different information retrieval phases to facilitate data mining (of original geodata) in general. All these different visualization techniques will be subject of user tests (as within the INSYDER project), comparing the usefulness and usability of the different approaches.

References

[1] Card, S.K., Mackinlay, J.D., Shneiderman, B. Readings in Information Visualization.

Using Vision to Think. Morgan Kaufmann Publishers, Inc, San Francisco, CA, 1999.

[2] Gershon, N., Eick, S.G. Information Visualization. IEEE Computer Graphics and Applications. July/August 1997, pp. 29-31.

[3] Göbel, S., Haist, J., Goebel, C. GeoCrystal: Graphic-Interactive Access to Geodata Archives. Proceedings SPIE 2002 – Visualization and Data Analysis, San Jose, CA, 2002.

[4] Göbel, S., Jasnoch, U. Metadata Information Systems for Geospatial Data, S.283-296 in Environmental Information Systems in Industry and Public Administration, edited by Claus Rautenstrauch, Susanne Patig. Idea Group Publishing, Hershey, USA, 2001. ISBN 1-930708- 02-5.

[5] ISO. Draft International Standard on Metadata for Geographic Information, ISO 19115- Metadata (status: DIS level), see http://www.statkart.no/isotc211/.

[6] Jasnoch, U., Göbel, S., Balfanz, D. InGeo IC: The portal to Geodata. Proceedings e2000, Madrid, Spain. Published in Standford-Smith, Kidd: E-business. Key Issues, Applications, Technologies. IOS Press, Amsterdam, 2000.

[7] Mann, T., Reiterer H.: A Combined Visualization Approach for WWW-Search Results in:

Proceedings of the IEEE Visualization 1999 (Vis ´99), San Francisco

[8] Rauber, A., Bina, H. A Metaphor Graphics based Representation of Digital Libraries on the World Wide Web: Using the libViewer to Make Metadata Visible. Proc. DEXA ´99, Florence, Italy, 1999.

[9] Reiterer, H., Mußler, G., Mann, T., Handschuh, S.: INSYDER – An Information Assistant for Business Intelligence, Proceedings ACM SIGIR 2000, Athens, ACM press, 2000, pp.112- 119.

[10] Robertson, B. biz viz gets real. Computer Graphics World, April 1999, S. 29-34.

[11] Spoerri, A. InfoCrystal: A Visual Tool for Information Retrieval. Proc. IEEE Visualization ´93, San Jose, CA, 1993, pp. 150-157.

[12] Väänänen, K. Metaphor-Based User Interfaces for Information Authoring, Visualization and Navigation in Multimedia Environments. Dissertation TU Darmstadt, 1995. Shaker Verlag, Aachen, 1995.

[13] Klein, P., H., Müller, F., Reiterer, H., Eibl, M. Visual Information Retrieval with the SuperTable + Scatterplot, accepted to the 6th International Conference INFORMATION VISUALISATION, 10-12 July 2002, London.

Referenzen

ÄHNLICHE DOKUMENTE

In most of the approaches proposed so far, the number of data items that can be visualized on the screen at the same time is quite limited (in the range of 100 to 1,000 data

Visualization of streaming data is strongly related to its temporal context and very often methods that map time to the horizontal axis are used to visualize the data stream?. How do

In constrast to this common tendency of process technology, we provide a straightforward, quickly viable alternative to IT-based process support at a reasonable effort: the

In addition to the direct involvement of the user, the main advantages of visual data exploration over automatic data mining techniques are: visual data exploration can easily deal

„[…] Daten (die für sich keinen Informationswert haben) werden zu Information (das heißt, es wird den Daten eine Bedeutung vermittels ihrer Organisation zugewiesen), die

Using temporal exploration, for instance, the user could identify an interval of high email traffic, and search in the spam folder within this interval for an email that was sent

The successful use of credentials for enforcing access control, and the consequent appli- cation of all the different trust management strategies that can be though of, requires

Accessing information over the Internet has become an essential requirement in modern economy, and unknown parties can come together on the Net and interact for the purpose of