Rocode >> Academic >> Publications

The following is a list of the published writings by Rohan Fletcher.

The highlight of his authoring career was the publication of his Masters research in Europe during 1996.

Rohan's theses are held in the Monash University library and you may use the links given below to visit the ACM archives of his conference publications.

Rohan has some knowledge of technical writing and typesetting. His theses and conference publications were written in the LaTex document formatting language. His poster was constructed using Microsoft Powerpoint.


Automated Production of Software Test Data

This thesis studies the testing of software and describes a language for the generation of test cases. Some testing methodologies are examined and their effectiveness for finding errors in software is discussed. Use of an automated testing environment should improve the confidence that software constructed is error free as potentially it will identify errors faster and more accurately than testing manually. The test generation specification (TGS) language is implemented in PC Scheme within an automated test generation environment called TGEN.

Go to the Monash Uni Library page for the thesis.

Thesis Abstract


Testing of Object Oriented Software Using Formal Specifications

This thesis examines the use of formal specifications in the testing of object oriented software. Formal specifications have the benefits of being precise and unambiguous. We investigate their role in developing methods for systematic test generation. We also give formal definitions of the test shell.

A framework is proposed for the joint development of software written in the object oriented language Eiffel, and its formal specification written in Object-Z. The framework gives rules for the development of the implementation of a system based on its formally specification, so that the implementation will interface with the test harness of the prototype system.

In an object-oriented environment, there is greater focus on the class as a test unit. Since a class has an associated set of attributes, object-oriented testing has a greater focus on object state. Our framework focuses on intra-class testing in a black box manner and does not deal with integration (or inter-class) testing of classes (cluster testing in the object oriented paradigm).

A testing system has been designed that will assist in the structuring, simple generation, re-generation, representation and execution of test cases based on the specification. The system is divided into several parts: the configuration subsystem, the generation subsystem, the system under test subsystem, the reference subsystem and the result verification subsystem. The prototype developed as part of this thesis demonstrates the main concepts of the system.

In short, the contributions of the thesis are as follows.

  • A formal definition of a test shell.
  • Procedures to create an intermediate test specification based on the formal specification.
  • Rules for the regeneration of test shells.
  • A methodology to create new test shells from an intermediate test specification.
  • The design and prototyping of a system to execute test shells using formal specifications in Object-Z and programs in Eiffel.

Go to the Monash Uni Library page for the thesis.

Thesis Abstract


A Framework for Testing Object Oriented Software Using Formal Specifications.

My first paper was published in Switzerland at the Ada-Europe'96 Conference.

Rohan Fletcher and A. S. M. Sajeev, Proceedings of International Conference on Reliable Software Technologies ( ADA-Europe'96 ), Lecture Notes in Computer Science, Springer-Verlag, New York, USA, 1996.

Published while a postgraduate of Monash University.

Go to the ACM page for the paper or visit Google Books.

Ada-Europe 96 Conference
Paper Abstract


Constructing Interoperable Components in Distributed Systems.

My second paper was jointly published in Australia at the TOOLS Pacific 1999 Conference.

Sea Ling, Heinz W. Schmidt and Rohan Fletcher, Proceedings of International Conference on Technology of Object-Oriented Languages and Systems (TOOLS Pacific '99), IEEE Computer Society, Los Alamitos, California, USA, 1999.

The separation of architecture definition language (ADL) and interface definition language (IDL) is widely accepted in the distributed system community. It separates issues of hierarchical composition and configuration from those of services, communication and synchronization and thus eases reconfiguration without changing the interfaces and implementation and allows the change of services within well-defined configuration constraints. Our approach towards distributed system definition draws on the Darwin project's notion of distributed system architecture definition and on Eiffel's notion of design-by-contract. We extend contracts to deal with synchronization and coordination at a higher level, explicitly targeting both large-scale loosely-coupled distributed components and tightly coupled parallel or multithreaded objects.In this paper, we present a Java binding of the resulting contract extensions. We incorporate annotations into Java programs adding Eiffel-style assertions and other contractual constraints. Java component code and constraints are then associated to architectural diagrams in a visual programming environment. The paper illustrates the language features in a case study of a lift control system. Based on the case study, the component-based design approach is illustrated and the internal synchronization of method execution is explained.

Published while employed at DSTC Pty Ltd jointly via The University of Queensland and Monash University.

Go to the ACM page for the paper.

Paper Abstract


Integrating Genomic and Clinical Data for Bioinformatics

My first poster was displayed in Australia at the Clinical Research Excellence '07 Conference as selected with 62 other scientific posters.

Fletcher, R., George, J., Thompson, N., Thomas, D., Bowtell, D., Fereday, S., Clinical Research Excellence 2007, Hawthorn, Victoria, Australia, 2007.

PeterMac is advancing the bioinformatics resources available to its researchers. In the age of genetically targeted treatments, reliable and rapid access to molecular and clinico-pathology details for in-house studies and published results is required. This is challenging due to the volume and complexity of the data involved. At PeterMac, the volume of genomic data produced was 151 GB in 2006 and projections for 2009 are as high as 17500GB. In addition, the variety of formatting and nomenclature used in different datasets necessitates a time consuming pre-processing step prior to analysis.

The Bioinformatics core has developed an easy-to-use web-based database, which not only stores and manages our genomic data, but is also linked it to clinical data. Clinico-pathology data from existing research databases are entered into the database via an ETL process. Clinical data from published datasets are also loaded and linked to its genomic data. This concept allows genomic and clinical data to be queried using both molecular and histo-clinical features. Researchers can request the resultant data formatted specifically for a number of analysis tools therefore, circumventing the need for data pre-processing.

This facilitates quick molecular profiling of histo-clinical features; or quick access to available histo-clinical measurements for given molecular imprint, thus improving the ease of more integrative analysis

Published while employed at Peter Mac Research via The University of Melbourne.

Go to the Peter Macallum Cancer Centre Research Report for January to December 2007. On page 135 are details of the Bioinformatics Core plus a record of my poster on August 18.