This thesis project is part of an exchange cooperation on "Formal Methods Protecting
Confidential Databases" funded by DAAD between Martin Giese at the Department
of Informatics of the University of Oslo, Norway, and P.H.Schmitt at the
Institute of Theoretical Computer Science, Universität Karlsruhe, Germany.
It is expected that the applicant will spend at least two months at the partner
institute in Oslo while working on this thesis. Funds covering travel and
living expenses are available.
Within the project "Formal Methods Protecting Confidential Databases", a formal
logical language will be developed that on one hand is powerful enough to express
the required queries that typically involve aggregations like summation, averaging and other
statistical operations and on the other hand can be mapped to large relational
As a first step towards this goal, the present thesis project will evaluate how
well the Object Constraint Language (OCL) and existing OCL query engines
meet these requirements. In particular the "iterate" operation of OCL is
important within this respect. Furthermore, a concept for coupling an OCL
query engine with a DBMS will be developed. In particular the question has to
be answered which functionalities of the query engine are better delegated to
the DBMS and how this can be done.
If time permits this concept will also be implemented for a suitable query
engine and an SQL database system.
The applicant should have solid knowledge on formal methods to the extent
covered by the lecture "Formale Systeme" at the University of Karlsruhe.
Basic knowledge of relational databases is indispensable. It will be neccessary
to aquire a detailed familiarity with OCL.
The applicant should enjoy teamwork and have an open mind for new perspectives.
A First-Order Query Language for Aggregation (at Uni Oslo)