+------------------------------------------------+
  |     OpenBQS - Bibliographic Query Service      |
  |  (PII - Platform Independent Implementation)   |
  +------------------------------------------------+

Authors:
    Martin Senger, Lichun Wang, Alan Robinson, EBI
    (senger@ebi.ac.uk)

=======================================================

Principles
----------

The OMG (Object Management Group, [OMG]) recently adopted a flexible
approach for specifying domain models in a technology independent way
(MDA - Model Driven Architecture, [MDA]). This approach is very
justifiable when designing language and technology independent models
and interfaces, but it is too broad for the implementation of such
interfaces.

Any implementation is written in a particular programming language
(Java in the case of OpenBQS - see below) - but using similar
principles as in MDA allows us to plug-in easily various components
based on different technologies. In particular, it makes it possible
to choose dynamically between:

 - using CORBA methods to access remote distributed components,
 - using RMI to achieve the same, or
 - using only local components, having everything in one address
   space.

We call this approach Platform Independent Implementation (PII, or
shortly a "pi"-approach).


Bibliographic Query Service - BQS
---------------------------------

Bibliographic search and citation are central to all scholarly and
research activities. Within the domain of life sciences research,
bibliographic citation is of particular importance for annotation of
large bodies of experimentally developed and computationally derived
data. 

This need was recognised by the LSR-DTF (Life Sciences Research Domain
Task Force, [LSR]) and a specification for such a service was recently
adopted [BQS]). The adopted standard represents an information model
[UML] of bibliographic citations and is based on the Dublin Core
Metadata for Resource Discovery [DC]. It includes also query
mechanisms to search bibliographic repositories and to retrieve
results.

OpenBQS
-------

OpenBQS [OBQS] is an open source implementation of the standard
Bibliography Query Service specification, using the PII approach
described above. Implemented in Java, it allows to choose between a
CORBA-based and local connection to a bibliographic repository. It is
being developed for accessing publication data from the EMBL
Nucleotide Sequence Database [EBI], and from the MEDLINE bibliographic
database. However, the PII allows to plug-in other repositories as
well.

Candy - the controlled vocabularies
-----------------------------------

Candy [CAND] provides a PII-based access to a set of controlled
vocabularies. It is a separate open source project used by OpenBQS to
introspect bibliographic repositories. It is a Java implementation of
another standard specification adopted recently by the LSR-DTF (as
part of the Genomic maps specification [MAPS]).

References
----------
[OMG]  http://www.omg.org
[MDA]  http://www.omg.org/mda
[LSR]  http://www/omg.org/homepages/lsr
[BQS]  http://www.omg.org/cgi-bin/doc?dtc/01-04-05
[UML]  http://industry.ebi.ac.uk/~senger/BQS/web
[DC]   http://dublincore.org
[OBQS] http://industry.ebi.ac.uk/openBQS
[EBI]  http://www.ebi.ac.uk
[CAND] http://industry.ebi.ac.uk/candy
[MAPS] http://www.omg.org/cgi-bin/doc?dtc/99-12-01