19th International CODATA Conference
Category: Poster, Interoperability
Keeping Biological Information Systems (BIS) Open and Simple
Peter T. Strobl, Mag. rer. nat. (peter.strobl@sbg.ac.at)
Arbeitsgemeinschaft fuer Vegetationsoekologie, Fachbereich fuer Organismische
Biologie und Oekologie, Paris Lodron Universitaet Salzburg, Austria
Biological Information System (BIS) is the name of a thoroughly
XML-based database management system for improving scientific workflow at the
Department for Organismic Biology and Ecology, University of Salzburg. When reengineering the
system from scratch, we payed special attention to adhere to open standards wherever
possible and remain platform independent in various ways. The system consolidates
several, previously independent database systems for taxonomy, syntaxonomy, and distribution data from different sources
in and around Austria. Even though BIS uses common interface
and data representation standards, we believe it is not a desirable aim to strive
for homogenizing the scientific database landscape using only one system type
or exchange path. In contrast, we decided to research on simplifying interoperability
of specialized biological systems. The atomic system entities of such networks
should not cover all possible functionalities, but only those which are really
needed, still maintaining the option to add or replace some modules at a later
point. Considerations on "return of investment" topics play a central
role in our system architecture; scarcely used components tend to pile up a
big amount of wasted development ressources. On the
other hand, the system should remain manageable and understandable by its target
user group, biologists with usually little professional information science
expertise. We do not consider graphical user interfaces to be the cure-all to
this challenge. Nowadays, when choosing technologies and implementation strategies
carefully, we believe it is possible to remove a lot of the complexity of open
standards for the user, and finally require only common computer operation experience
to deal with the system. Examples for this approach will be given. Because of
the above outlined topics the main goal for biological information systems of
our type shall be to keep things simple as possible. BIS' design rules enable access to information
system technologies even to low-budget research, whether in "developing
countries" or in the "first world". To these scientists we freely
offer BIS know-how to promote scientific workflow
and networking in new ways and to make a small contribution for freeing scientific
work from financial obstacles. The author is looking forward to having open
discussions on this and other related subjects.