Making the Environmental Data Catalogue UDK and other Databases Available on the World-Wide-Web

Ralf Kramer, Forschungszentrum Informatik (FZI), Karlsruhe
Horst Spandl, Landesanstalt für Umweltschutz (LfU), Stuttgart

Home | Index Page | Previous | Next

  1. Introduction
  2. Accessing Databases from the World-Wide Web
    1. World-Wide Web Basics
    2. World-Wide Web Access to UDK
  3. Beyond the World-Wide Web
    1. Motivation
    2. Using te Common Object Request Broker Architecture
    3. Using JAVA
  4. Concluding Remarks
  5. References

1 Introduction

Large scale environmental information systems, especially those operated by public organisations, are usually based on a variety of data sources. The underlying systems are heterogeneous both on a technical level (e.g., different operating systems like VMS, UNIX and PCs running Windows; different database systems, e.g., relational database systems, pre-relational one, and geographic information systems) and on a semantic level, i.e., heterogeneous with respect to their content. Some of the relevant information may not even be available online at all. Due to the diversity of the implemented systems and their respective user interfaces the art of retrieving the required information isrestricted to a fairly small group of users that have the necessary technology and knowledge at their hands.

As an example for such a system, Figure 1 illustrates the concept of the environmental information system UIS Baden-Wuerttemberg, Germany [Mayer-Föll 93]. In a three layer architecture, systems are organized hierarchically and vertically.

The remainder of this paper is organized as follows. In Section 2 we focus on making databases available on the World-Wide Web using the current standard access technique, the so-called common gateway interface CGI. In Section 3 , we show how World-Wide Web techniques can be combined with further standards and hence, how advanced functionalities can be provided. Section 4 concludes the paper with some final remarks.

Figure 1: The Component Categories within the Environmental Information System (UIS), Baden-Württemberg, Germany

2 Accessing Databases from the World-Wide Web

2.1 World-Wide Web Basics

In recent years, the Internet and especially the World-Wide Web (WWW) has changed the situation described in Section 1 dramatically. One of the main attractions of WWW technology is that it offers a common platform-independent user interface to access potentially distributed information sources for no or fairly little extra cost. This provides the tempting opportunity for any information provider to serve a larger audience. At the same time the usual road block of interface development is reduced to serving a simple standard output protocol, the Hypertext Mark-up Language (HTML). The presentation is delegated to the individual web browser. WWW browsers like Mosaic and Netscape are available on all major platforms. Besides making staticWWW pages in the standard language HTML available on the web, WWW servers as well allow for accessing arbitary data sources, e.g., database systems, via the common gateway interface (CGI). Hence, all databases that are already in place are potentially available on the web as well. All that is required is to code a CGI program to access the database which generates HTML as its output. Furthermore, the reader should bear in mind that WWW technology can be used both on the Internet and on private networks, so-called Intranets, as well.

An important issue in the Word-Wide Web as in any hypermedia system is to identify and to find relevant data sources that are likely to answer a certain query. Within the Web, various search engines like Yahoo, Lycos or Altavista have taken on the task to provide a searchable index of available HTML documents. Unfortunately, these systems can only catalogue static HTML pages.Dynamically generated HTML, which is typical for database access through the CGI interface, is beyond the capabilities of these indexing programs, so-called web-crawlers or web-robots. An additional difficulty creeps in from the semantic dimension of the problem as most web-robots rely on information available in the indexed HTML document itself, typically the <TITLE> of the document in the header section. The quality of the descriptions depends on the individual effort of the information providers. However, it is impossible to apply a common glossary throughout the world!

2.2 World-Wide Web Access to UDK

Within the Ministry of Environment Baden-Wuerttemberg the enviromental data catalogue Umweltdatenkatalog (UDK) [Swoboda 95, Lessing 9] is used to collect meta-information on data sources that are relevant to the environment. The UDK is a joint development effort of the German Federal Ministry of Environment, most of the German states and Austria. One major aspect is the categorisation of the stored informationon the basis of an integrated special environmental glossary, the Umweltthesaurus (environmental thesaurus) of the German Umweltbundesamt which opens the door to more sophisticated search techniques. Consequently, taking into account that WWW technology can be used on private networks as well, accessing environmental data catalogues using the WWW and its tools is a promising approach for both Intranets and the Internet.

At FZI, work in this area started as early as 1994 [Kramer 95, Kramer 96a] under contract of the Ministry of Environment Baden-Wuerttemberg, Germany. The environmental data catalogue UDK has been made available as an online database via the common gateway interface. The Austrian version, which is already available to the public (URL: http://udk.bmu.gv.at), features both German and English query and result forms. A special feature of the version for Baden-Wörttemberg, Germany, is the online access to underlying data sources like further databases and environmental reports. Furthermore, access from and to a geographic information system is available in a first prototype [Koschel 96, Kramer 96b]. A deliberate attempt has been made to add the additional functionality with the least extensions to the underlying UDK data model as possible. All of these databases have been made available using the standard common gateway interface at the WWW server site.

3 Beyond the World-Wide Web

3.1 Motivation

WWW is an attractive technology to provide cost-effective access to information residing in database for a wide audience. Especially for simple read-only accesses to few relations in relational database systems, the current standard WWW technology as presented in Section 2 provides adequate solutions.

On the other hand, when compared to classical client/server database tools, WWW poses some problems. One of these problems is that WWW only supports access to a single data source at a time, i.e., if, in order to answer a certain query, several data servers have to be accessed, the user has to access all of them in turn explicitely. He or she has to combine the individual results, which are presented on individual result pages, manually. Another severe limitation is that the hypertext transfer protocoll (HTTP) does not support more complex interactions with database systems, so-called database transactions. Hence, especially database transactions that access and modify more than one database relation are extremely difficult, if at all, to implement. The connection-less communication model of HTTP (re-)imposes restrictions on the user interface. Common techniques such as context-sensitive list of values or pop-up menus are currently out of reach of the current WWW standards.

Figure 2: A Sotware Architecture incorporating the World-Wide Web and COBRA

3.2 Using te Common Object Request Broker Architecture

The first of these problems, namely accessing several data sources simultaneously using WWW technology, can be solved by introducing an additional so-called middleware layer. This middleware layer can be implemented using the common object request broker architecture [Ben-Natan 1995],a standard that is supported by more than 600 companies in the Object Management Group OMG [Object Managment Group 95] (URL: www.omg.org). Based on an evaluation of the available CORBA implementations [Koschel 95] the architecture of Figure 2 has been developed [Koschel 95a, Koschel 96].

Horizontally, we distinguish between application level and system level services.System level services provide basic functionality such as accessing databases and preparing HTML pages. Application level services basically combine several system level services into higher level services that are available for the user. Using CORBA, these services can be transparently assigned to computers in a network.

Vertically, we distinguish between information and data services. Information services help the user to find relevant data sources, e.g., data services and environmental reports. They are based on the environmental data catalogue UDK, which has been extended in order to provide online access to data services as described in Section 2.2 and shown in Figure 2. These data sources primarily are environmental databases and reports that are as well accessible with WWW tools. Data sources may comprise geographic information systems (GIS), relational, pre-relational, object-oriented database systems, and expert systems.

3.3 Using JAVA

The second problem, namely complex database interactions across the net, can be solved by using a new programming language, JAVA, in so-called applets, i.e., small programs that are executed within WWW browsers. These applets communicate with server processes using more powerful communication protocolls than HTTP. Currently, commercial implementations that combine both CORBA and JAVA become available in first releases. The very first one available, PostModern Computing's Black Widow [Computing 96), has already been tested successfully at FZI.

The third problem of using more context-sensitive presentation techniques within the user interface can be approached using JAVA or JAVAScript, a scripting extension to HTML introduced by the latest Netscape versions. Unfortunately, in sharp contrast to JAVA, JAVAScript still lacks important security features and the strict definition as a standard. Hence, at least currently, using JAVAScript cannot be recommended.

4 Concluding Remarks

In this paper, we outlined several approaches for accessing databases from the World-Wide Web. WWW-UDK 3.0 [Kramer 95, Kramer 96a, Kramer 96b], a tool developed at FZI that makes the relational UDK database available on the WWW, is currently used by three ministries of the environment in Austria and Germany. The deficiencies of standard WWW technology when accessing databases have been addressed at FZI employing further widely accepted standards, namely CORBA and JAVA. These three technologies taken together, World-Wide Web, CORBA, and JAVA, have the potential to change our current computing environments dramatically.

Acknowledgements: The first author would like to thank his colleagues at FZI, Arne Koschel and Ralf Nikolai, for the stimulating discussions and inspiring work atmosphere, Rudolf Legat and Dr. Konrad Zirm at the Austrian Federal Ministry of Envrironment, and Bernhard Lorenz at ISEP, Vienna, Austria. It has been a pleasure to collaborate with Manfred Mueller and his colleagues at the Landesanstalt fuer Umweltschutz Baden-Wuerttemberg, Karlsruhe, Germany, and with Dr. Joachim Wiesel and Wilhelm Hagg, Institute of Photogrammetry and Remote Sensing, University of Karlsruhe, Germany.

References

[Ben-Natan 95] R.Ben-Natan. CORBA A Guide to the Common Object Request Broke Architecture. J. Ranade Workstation Series. McGraw-Hill, New York, USA, 1995.

[Computing 96] PostModern Computing. Blackwidow information. URL http://www.pomoco.com/BW/reference_manual.html, 1996.

[Huber-Wäschle 95] F.Huber-Wäschle, H. Schauer, P. Widmayer, editors. Herausforderungen eines globalen Informationsverbundes für die Informatik; 25. GI-Jahrestagung und 13. Schweizer Informatikertag, Zürich, 18.-20. September 1995/GISI 95, Informatik Aktuell. Springer, 1995.

[Koschel 95a] A Koschel, R. Kramer, R. Nikolai. Architektur des WWW- und CORBA-basierten UIS. In R. Mayer-Föll, A. Jäschke, editors, Projekt GLOBUS; Konzeption und prototypische Realisierung einer aktiven Auskunftskomponente för globale Umwelt-Sachdaten im Umweltinformationssystems Baden-Wörttemberg; Phase II 1995, number FZKA 5700 in Wissenschaftliche Berichte, pages 39-45. Forschungszentrum Karlsruhe Technik und Umwelt, Karlsruhe, Dec. 1995.

[Koschel 95b] A. Koschel, R. Kramer, D. Theobald. CORBA-Evaluierung för das UIS Baden-Wörttemberg}. In R. Mayer-Föll, A. Jäschke, editors, Projekt GLOBUS; Konzeption und prototypische Realisierung einer aktiven Auskunftskomponente für globale Umwelt-Sachdaten im Umweltinformationssystems Baden-Wörttemberg; Phase II 1995, number FZKA 5700 in Wissenschaftliche Berichte, pages 47-92. Forschungszentrum Karlsruhe Technik und Umwelt, Karlsruhe, Dec. 1995.

[Koschel 96] A. Koschel, R. Kramer, R. Nikolai, W. Hagg, J. Wiesel. A Federation Architecture for an Environmental Information System incorporating GIS, the World-Wide Web, and CORBA.In Third International Conference/Workshop Integrating GIS and Environmental Modeling, Santa Fe, New Mexico, USA, Jan. 1996.

[Kramer 95] R. Kramer, H. Spandl. Metadatenzugriff in Weitverkehrsnetzen: Eine Realisierung am Beispiel des Umweltdatenkatalogs UDK. In Huber-Wäschle etal. [Huber-Wäschle 95], pages 610-617.

[Kramer 96a] R. Kramer, T. Quellenberg. Global Access to Environmental Information. In R.Denzer, D.Russel, G.Schimak, editors, Environmental Software Systems; Proceedings of the International Symposium on Environmental Software Systems, 1995, International Federation for Information Processing (IFIP), pages 209 - 218, London, 1996. Chapman and Hall.

[Kramer 96b] Ralf Kramer, Ralf Nikolai, Andree Keitel, Rudolf Legat, Konrad Zirm. Enhancing the Environmental Data Catalogue UDK for the World Wide Web. In 10. Symposium Informatik für den Umweltschutz, Sept. 1996. to appear.

[Lessing 95] H. Lessing, O. Gönther, W. Swoboda. Ein objektorientiertes Klassenkonzept för den Umweltdatankatalog (UDK). In H. Kremers, W. Pillmann, editors, Raum und Zeit in Umweltinformationssystemen; Space and Time in Environmental Information Systems; 9th International Symposium Computer Science for Environmental Protection CSEP'95, number~7 in Umwelt-Informatik Aktuell, pages 391-399, Marburg, 1995. Metropolis.

[Mayer-Föll 93] R. Mayer-Föll. Das Umweltinformationssystems Baden-Wörttemberg - Zielsetzung und Stand der Realisierung. In A. Jäschke, T. Kämpke, B. Page, F.J. Radermacher, editors, Informatik för den Umweltschutz, Proceedings 7. Symposium GI-FA 4.6, pages 313-337, Berlin Heidelberg, 1993. Springer.

[Object Management Group 95] Object Management Group. The Common Object Request Broker: Architecture and Specification, Version 2.0. (OMG) Document, Object Management Group, Inc. (OMG), July 1995.

[Swoboda 95] W. Swoboda, H. Lessing, P. Grolimund, O. Gönther, U. Haas, R. Legat, M. Vogler, K. Zirm. Metadatenklassen im Umweltdatankatalog (UDK). In Huber-Wäschle etal. [Huber-Wäschle 95], pages 601-609.

Home | Index Page | Previous | Next