ESSE: Development of an Expert System for Software Evaluation
The main aim of this project is to develop an expert system for software
evaluation. Software evaluation is the procedure that compares a set of
software products based on several criteria. It is evident that software
evaluation is particularly difficult and many different and contradictory
variables at various intermediate steps must be considered and combined
in order to reach a final decision. Due to the uncertainty of the problem,
it is really hard to be carried out manually, especially from non expert
people. The ESSE (Expert System for Software Evaluation) project aims to
the design and development of an Expert System to evaluate the combination
of the attributes of software products. ESSE takes into account several
software attributes, such as quality, cost, time, adherence to standards
e.t.c. In addition, a well known and understood operational research method,
namely the multiple criteria decision aid (MCDA) is employed to cope with
the variety of attributes that affect software evaluation. ESSE elaborates
the theoretical results and practical data of previous estimations to fine
tune the process of software evaluation in the future.
The ESSE System is being developed in a PC-486 platform. We will use
the LPA Flex Expert System Shell. Flex interacts with the underlying language
LPA Prolog for MS-Windows, so the final ESSE product will be a combination
of these two platforms.
Flex will be used for the construction of the ESSE rules, which will
be used for the definition of the problem and for the search of the solutions.
LPA Prolog will be used for the interface between the ESSE System and the
end user. LPA Prolog supports all the MS-Windows control mechanisms, such
as text, list and combo boxes, menus e.t.c.
ESSE Architecture comprises of three main parts, the Kernel, the Knowledge
Base and the Intelligent Front End.
The IFE has a role to aid the user in collecting the necessary information
and in validating the model created. Validation will be provided both through
tutorials (on line help) and through verification procedures.
The knowledge base consists of three parts. These are the rules, the frames
and the historical data. Rules are used mainly in the application running
step by the inference engine. Frames have to be defined for each type of
objects used by the ESSE machine. Historical data are the instances of
the frames described above. They have the same meaning and use like the
objects in other object-oriented environments.
The kernel of ESSE machine consists of the Flex inference engine and the
Assist. professor, Dept. of Informatics (coordinator)
Ioanis Stamelos, Lecturer, Dept.
Petros Kefalas, Assist. professor,
Senior Researcher at LAMSADE, France
18 months (1/6/1996-30/11/1997)
General Secretariat of Research & Technology, Ministry of Development,
Project Programme PENED 1996 (PENED Project No. 1275 - AUTH Research Committee
Project No. 1696)
Department of Informatics Home Page
LPIS Group Home Page
For any comments regarding this page contact Nick