Project proposals
for MSc. students.
Get in touch if
you are interested.
mail suggestions to
omer rana
Note:
These projects are
unoffical and aimed
mainly at determining
interest in the
outlined area. Dr. Lawson
will be officially
offering projects after your
examinations in the summer
term.
| These are project proposals for 1997-1998. New project proposals for 1998-1999 will be added in mid April 1999. |
|---|
Proposal 1: Integrity Constraints
-
To design a Java application, using Visual Age for Java and
JDBC drivers to connect to the departments Oracle
server.
- Analyse the current system, and look at ways to improve the functionality
- Analyse the codasyl model and produce an accurate relational model.
- Analyse the relational model, to determine the kinds of integrity constraints that should exist.
- Design and implement the schema in Oracle on Solaris.
- Design, implement and document a new version of the CAS system using Visual Age (IBM) for Java, Visual J++ (Microsoft) or JBuilder (Borland)
- Port this Java application to Solaris and test using the Sun Java station.
-
To migrate a CODASYL legacy database system to a JAVA database application, using
JDBC and ORACLE.
The legacy system is the Course Administration System used by the Continuing
Education department.
This project will be divided into a number of sections:
This project is large enough to support 3 MSc students working on different parts of the system.
The primary aim of this project is to look at what kinds of integrity constraints exist in legacy systems, and to determine what integrity constraints should exist.
This project will build on work carried out by Nic Caine as part of Ph.D. research. There are various modules that may be of relevance in the project, and the candidate will be encouraged to consider multiple modules. There is scope for publication of this work, and possible inclusion in a future research report.
If students are serious about undertaking this project, then please contact Nic Caine scmnjc@cs for further information about the project.
![[to top of page]](top.gif)
Proposal 2: Java Security
-
To investigate the deficiencies in the sandbox through
which Java applets interact with applications.
The candidate should be able to understand the underlying principles of the Java Virtual Machine.
A large number of interesting applications are possible in areas like Electronic and Internet Commerce for using the techniques learnt in this project.
Proposal 3: Resource management in Distributed Systems
-
Jointly with Prof. David Walker
To investigate the use of resource management policies in a distributed computing environment using CORBA. The project will require the analysis of object granularities and their effect on application performance.
The interested candidate will learn about CORBA and the general aspect of network and system management. The aim of the project would be develop an overall framework which could then be incorporated into a future research project.
Proposal 4: Neural Network simulator in Java
-
Jointly with Prof. Antonia Jones
To develop a neural network simulator in the Java programming language, where the different neural modules could be spread across multiple machines. The interested candidate will be given help with developing neural network models, but will have to work towards identifying how these modules could be made to interact in a distributed environment using Java.
Proposal 5: Graphical visualisation of data using Java 3D API
-
Jointly with Sun Microsystems (Bristol)
To develop a means of visualising numeric and symbolic data from
SQL queries using the new Java 3D API (scheduled for release in March).
You will learn about extracting data from a database such as Oracle
and visualising it via a Java applet or application.
Alternatively, candidates may also use the Java 3D API for
visualisation from programs such as neural networks.
Please ask for details if you are interested in this project proposal.
Proposal 7: A financial application using Mobile Agent (in Java) libraries
The candidate will be required to develop a simple financial application using a mobile agent library provided. The aim of the project is to understand how to use agent techniques and libraries for a particular application, and to investigate the issues involved in dealing with mobile code.
We currently have two libraries available, and therefore two candidates may do this project using one of the two libraries.
Proposal 8: A personal Web Agent (in Java)
The aim of the project is to understand issues involved with designing an agent which can monitor user behaviour when accessing web pages. The agent should build a personal profile of the user, and use it to customise queries to standard search engines such as Alta Vista or Lycos.
The first part of the project is to investigate a personal web agent such as WebMate (which will be provided to the candidate along with documentation), and then to consider techniques for improving its functionality and (or) use artificial intelligence techniques based on rules, a neural network based technique or simple statistical forecasting techniques.
The ideal candidate should have done either the evolutionary computing or the AI course. Alternately, if you have some mathematical flair - you will find this an interesting project.