This page is now obsolete. Please go to http://www.joehalliwell.com/ instead.
My name is Joe Halliwell, which is an anagram of Elijah Lowell. I'm a PhD student in the Division of Informatics at the University of Edinburgh. I'm a member of the Approximate and Qualitative Reasoning group which is part of the Center for Intelligent Systems and Applications. My supervisors are Qiang Shen, and Alan Smaill.
I enjoy walking around in an aloof manner, drinking beer, eating food, sleeping all day, coding in Python, reading books, talking to people and smoking cigarettes. My e-mail address is firstname.lastname@example.org. This is what I look like.
I help to run a community website dedicated to constrained literature.
I am working in the general area of representing and reasoning with imprecise information. I am especially interested in the relationship between fuzzy and stochastic uncertainties and hybrids of the two.
The present: I have developed a theory of approximate linguistic probability assessments (such as "It is very unlikely to rain"). This is strictly weaker than the classical theory (in that certain theorems do not carry over) but extends the classical model (in that classical probability measures are a special case of linguistic probability measures.) This work has been presented recently at WCCI'02. I am currently developing a suite of graphical tools to facilitate the design, editing and querying of networks of linguistic probabilities.
The future: I will address two questions related to the semantics of linguistic probabilities. First, what epistemic processes give rise to (fuzzy) uncertainty about the probability of a given event? And second, what are the decision-theoretic (or behavioural) implications of linguistic probabilities?
The outcome: In answering these questions I hope to discover how linguistic probabilities might enhance an agent's decision-making capabilities and demonstrate these advantages in a simple multi-agent trading system.
J. Halliwell, J. Keppens and Q. Shen, Linguistic Bayesian Networks for Reasoning with Subjective Probabilities in Forensic Statistics, Proceedings of the 9th International Conference on AI and Law, 2003
J. Halliwell and Q. Shen, Towards a Linguistic Probability Theory, Proceedings of the 11th International Conference on Fuzzy Sets and Systems (FUZZ-IEEE '02), 2002.
J. Halliwell and Q. Shen, Towards Temporal Linguistic Variables, Proceedings of the 10th International Conference on Fuzzy Sets and Systems (FUZZ-IEEE '01), 2001.
J. Halliwell and Q. Shen, From fuzzy probabilities to linguistic probability theory, Proceedings of the 2001 UK Workshop on Computational Intelligence, pp. 129-135, 2001.
J. Halliwell and Q. Shen, Towards fuzzy temporal logic: Modelling dynamic systems, Proceedings of the 7th UK Workshop on Fuzzy Systems, pages 126-136, 2000.
I hope to make electronic versions of some of my publications available soon, but in the meantime please e-mail me for copies.
- I'm tutoring the Third Year Large Practical (3LPCS). Tutorials take place on Mondays at 1600 in room 4.03.
I'm mainly working on Arbor which, it turns out, is a framework for writing programs that operate on graphs. It's written in Python (using the Tkinter bindings for Tk for the GUI). Here's a screenshot.
Arbor is kindly hosted in CVS by the nice folk at Twisted Matrix Laboratories. It's currently in a pre-alpha state, but bleeding-edge (AKA broken) sources are obtainable by issuing the following commands:
cvs -d$cvsr login
cvs -d$cvsr checkout Arbor
I've hacked together my own window manager called
rawm which stands for Really Ascetic Window
Manager (screenshot, source tarball)
Also: a python script for xchat supporting nick maintenance and linkbot colourisation (screenshot, source); a theme for the Enlightenment window manager called E-legence (screenshot, tarball); a stupid ALife thing prompted by an argument about the rationality of altruism (screenshot, another screenshot, source tarball); some scripts for managing CDs full of MP3s (source tarball); SmartFolders that improve Zope's FTP support (tarball); and some other stuff I've forgotten about probably.
I've made some things, words, sounds and pictures
This website was made with XEmacs, gozer, Portable Network Graphics, OpenSSH, The Gimp, Debian GNU/Linux, zsh, Galeon and Python. I use these tools every day and they're great. The font used for the headers is called The Perfect Drug.
Last modified: Sun Jun 16 21:26:01 BST 2002 Comments? Questions? Please e-mail email@example.com.