Map · News
Home » Teaching »
Project Ideas

../photo/pcd0423/016.php

A thesis project should involve some measure of originality and critical thinking; it is more than just a hack. Below are a few general ideas for projects that continue to interest me.

I am happy to advise well-qualified LIU students on thesis projects in these and related areas, subject to these restrictions:

  1. You must provide me with a resume, transcript, and a recommendation from another CS faculty member.
  2. You must write a 4 to 6 page proposal of what you intend to do before I will sign your registration form. This means you should start planning a month or more before your thesis semester starts.
These restrictions help to ensure that we do not waste one another’s time. A thesis should be a thoroughly challenging and rewarding experience.

I welcome feedback, interest, and ideas from folks outside LIU as well!

../photo/pcd0091/047.php

Programming languages for web publishing

 More on this soon…
../photo/pcd0077/022.php

Database-backed email system

 Electronic mail clients use several different methods for storing messages so that they can be retrieved via browsing and searching. Typically, there is a concept of a folder, which is implemented either as a single file (with messages concatenated) or a directory in the underlying file system (with each message in a separate file). Each method has its own scalability issues and searching limitations. Messages seldom can belong to more than one folder. Two messages that belong to the same thread, but end up in different folders (perhaps because they are from different senders) are unlikely ever to be seen together again. And in many systems, a mis-filed message is as good as lost.

I believe that using a relational database to store messages would allow us to move beyond the folder and file-system limitations of current email systems. Note that this project is mainly about extending the user interfaces and protocols of email software. Why bother substituting a database for the simpler files and folders if it doesn’t expand the capabilities or expressiveness of the system?

I did some preliminary work on this a few years ago: please read this report (the Perl scripts it mentions are in this zip file). I would expect a thesis project to go a few steps further, not just to repeat what I’ve already done.

Published July 2003
Updated 25 July 2003