BUKU ABSTRAK - Universiti Putra Malaysia

Soft Systems Methodology for Environment-aware Case-based Reasoning Systems
Dr. Sa’adah Hassan
Faculty of Computer Science and Information Technology, University Putra Malaysia,
43400 UPM Serdang, Selangor, Malaysia.
+603-8946 6507; saadah@fsktm.upm.edu.my
Keywords: Case-based reasoning, soft systems methodology
Real-time Audio Transcription Technique using Static and Dynamic Window Sizes
with a Low-cost Hardware Interface
Dr. Shyamala C. Doraisamy
Farshad Arvin
Faculty of Computer Science and Information Technology, University Putra Malaysia,
43400 UPM Serdang, Selangor, Malaysia.
+603-8946 6548; shyamala@fsktm.upm.edu.my
Keywords: Audio transcription, signal processing, microcontroller
273
Science, Technology & Engineering
A Case-Based Reasoning (CBR) system solves new problems by remembering a previous similar problem
and reusing its solution. Descriptions of previous problems and their solutions are stored in a case base. The
system depends on information given by the end-user and historical cases for the descriptions of a new problem
and its solution. Thus, the ability of the system to make appropriate decisions or suggestions is limited. While
the characteristics of today’s computing systems offer a lot of opportunities to be explored, it is beneficial for a
computer system to have an awareness of its environment by providing information on the current situation to the
end-user and to the computer system itself. Such contextual information is useful to keep the system informed,
define the problem descriptions, and recommend appropriate solutions or suggestions based on the current
situation. CBR systems with environment awareness can potentially improve to define the problem description
and its solutions precisely with the contextual information. However, the development of a system that has
environment awareness requires a deeper analysis of the system itself and its relationships with its environment.
The analysis approach in the systems development should be able to yield a collection of collaborating elements
in a computing system and its links to the environment. Today’s computer systems are situated in intricate
environments, and are required to adapt to the dynamic environment as well as be proactive in making decisions.
Such requirements are difficult to be modelled using the ‘hard’ analysis approach. The Soft Systems Methodology
(SSM) is an established approach to understand an ill-defined problem in such situations. This research presents
the SSM analysis in the development of an environment-aware CBR system. The analysis models help clarify and
define the system’s environment explicitly to describe the design of the system.
A real-time signal processing technique using static and dynamic window sizes with a low-cost hardware
interface is proposed. Most audio signal processing systems require high cost hardware interfaces such as DSP.
The proposed technique in this project introduces a hardware interface that includes a microcontroller to process
audio music signals for pitch extraction. The audio signal is divided to smaller sections known as windows to
obtain samples of the signals for transcription. In general, two different approaches are proposed which convert
the voice samples for real-time processing using static and dynamic window sizes. However, the transcription
process into MIDI data involves complex calculations and in this proposed technique, a simple technique is used
to process audio signals. The transcribed data generated shows the feasibility of using microcontrollers as a realtime
MIDI generation hardware interface.