Information Management in Natural Science:
PhysNet is a worldwide Network for Physics Information.
Information on Departments and Documents
are offered in this non-commercial, professional, free-to-use, and
globally distributed network. PhysNet is a portal to Physics-related
information on Journals, Conferences, Jobs, Educational Resources as well as
administrational information and grey literature, which are publications offered on the
authors own institution's server.
The talk will give an overview on the latest developments in using metadata in the field of documents as well as on-textual objects (experts database). Also an overview will be given, on how to build up distributed information services.
MareNet is an online information service for marine
science under the auspices of the German Society for Marine Research,
developed and provided by the Institute for Science Networking at the
C.v.O. University of Oldenburg.
MareNet is a so called distributed information service which makes pragmatic use of the advantages of the web. All the information served by MareNet is gathered from the web servers of worldwide distributed marine research institutions, departments and institutes of universities and various related service providers. The restriction to those professional institutions ensures the quality and relevance of the offered information. All information of MareNet is kept, stored and maintained by its creators at their local institutions server and the creators retain all rights of their data.
MareNet is one of the largest and completest Portals for Marine Science.
It is a noncommercial, professional, free to use, globally distributed
up-to-date online information service which enables marine scientists to
keep in touch with
the worldwide marine science community and to receive all
specific information they may need.
MareNet provides a set of services, like list of links to more than 500 marine research institutions, list of links to scientific publications and other document sources of this institutions, lists of online available journals, dataservices, lists of various releted jobs and conference sites and links to other related information services.
On the one hand, dissertations are the most important documents of
personal qualification for young scientists. On the other hand,
dissertations also represent a significant part of the current status of
scientific research in the literature. They include recent research
findings and innovative methodical approaches and are therefore
especially important for the scientific public in the field. Here
immediate, current accessibility is critical.
In view of the different demands posed on dissertations by societies and scientists in any field, global access to the documents from a single source remains impossible. A foundation for the handling of electronic dissertations in Germany has been laid in the project Dissertations Online. However, in the field of mathematics, additional points stemming from the relative frequent inclusion of formulas in mathematic dissertations must be considered.
Within the scope of the project MathDiss International, a permanent international online full-text document server for mathematical dissertations will be established.
In this connection, questions concerning an online presentation of the documents and the problems of long-term archiving (from TeX resp. LaTeX documents) will be considered. They include the question of how to homogenize such files in order to enable their later conversion into programming languages following XML. Furthermore, the expansion of research possibilities using online documents is being planned. Providing access to the tables of contents, lists of tables and illustrations and bibliographies on the LaTeX level is of top priority. Because of the structure of mathematical documents written in LaTeX we have a lot of high quality information which gathers dust in the archives without being used for the retrieval of scientific documents. This situation should be changed and it could be changed because LaTeX has become a widely accepted tool in mathematical literature.
The focus of CARMEN (Content Analysis, Retrieval and Metadata: Effective Networking ) is on
content analysis. New tools in the subject with strong connection to
Efforts to give homogeneity and consistency to decentralized information environments try to create information systems for distributed data repositories. Quite often the approach is by technical means to physically gain access to differently structured spaces of documents. This will not suffice: The main problem of different content type and conceptional design of the various data collections is not solved this way.
The project therefore approaches content analysis with developments and prototypical implementations in three fields:
Math-Net as an information portal for mathematics offers a couple of different sevices such as searching for, and in, documents, personal and adminstrative data, software and links. Have a look at the site www.math-net.de. PhysNet also offers a huge amount of information from physicists for both physicists and non-physicists, see www.physnet.de. Here you can also find job offers and educational material.
If you compare the two entry web pages you notice the differences:
There are two expanded services with different interfaces optimized
for specific users. The Learned Societies uses their own classification
leading to other retrieval methods.
Both services have in common is the underlying structure: decentralized servers
that are building a network; and being indexed by a shared software tool called
If you compare mathematics and physics thematically you find them really similar to natural sciences. They both describe things in a very formal way, so publications consists of plain text, formula and graphics. In physics there may be additional experimental data and construction plans or software allgorithm.
Keyword and classification schemes are done by MSC and PACS.
In doing research on similar topics, the groups often use only one scheme: One example: In mathematics a special construction leads to the 'Hilbert space' in 'body theory'. This is used in physics, too in 'quantum mechanics' and 'field theory'.
For interdisciplinary searching both nominations may be useful.
Homogenity is not easy to reach. First, the author can use both schemes.
But later you need correlations within the classification schemes.
One part of the German research project CARMEN is trying to set this up.
You have this problem again in telecommunication. Here a third player, computer science and engineering with its schems AMC is involved. Up to now, there is no portal or common scheme set up.
feel free to contribute, please contact Kerstin Zimmermann |
upload 15.4.2001, update 24.8.2001