Profile

My research falls in an area that is called artificial intelligence. AI is part understanding what is intelligence, how can we recognise it and part providing programs/systems with capabilities to think and act without constant human intervention. I am interested in the latter, in making system autonomous. At the same time I am interested in what computer scientist call complex problems: problems for which solutions are easy to verify (that they are indeed solutions) but that are difficult to find.

A lot of computer science involves mathematics and I really like that aspect. So while some computer scientists are more drawn to the human interaction/psychology side of the science, I prefer working with tools that can be mathematically proven.

For those of you that have done programming before, you probably have worked with languages like phd, python, java, visual basic. These are what we CS people call procedural or object oriented programming languages. As a programmer, you have to tell the program step by step how to solve the problem. Once you have done that, you can run the program and the program will give you the solution.

But have you ever come across logic programming languages? Prolog? These are so-called declarative language. Rather than stating how to solve a problem, you describe the problem and how a solution needs to look like. Once you run the program, you will get the solution to the problem. Declarative programming languages are the type of languages that I am interested in. They allow you to relatively easy solve a problem without having to know how to solve it.  For those interested in mathematics, these languages have a very nice mathematical foundation.

I use this programming language to solve a variety of problems. One example is ANTON, an automated music composition system: a program that composes pieces of music with me only telling it how long the piece should be and whether it should be a solo or a duet. shows a piece Anton wrote as my wedding gift. An other application is PERPETUATE. It is used to check how damaged buildings are after an earthquake and how they might collapse. shows one of the collapse mechanism where one side of the building detached due to severe vertical cracks on the side.

Since you (hopefully) will be asking me questions, I have a question for you. Can you guess what this program does? Do not worry too much about the format of the program. Instead read the comments (lines starting with %%).

%% There are 9 possible locations in each of two dimensions
position(1 .. 9).

%% Each square may take a value of 1 to 9
value(1 .. 9).

%% state(X,Y,N) is true if the state of square X in row Y has value N

%% Each square may take any value
1 { state(X,Y,N) : value(N) }  1 :- position(X), position(Y).\

%% A number cannot appear twice in a column
:- state(XA,Y,N), state(XB,Y,N), position(XA), position(XB),
position(Y), value(N), XA != XB.\

%% A twice in a row is not allowed
:- state(X,YA,N), state(X,YB,N), position(X), position(YA) ,
position(YB),value(N), YA != YB.

%% Define the concept of locality
%% These are 3 by 3 sub squares
sameSubSquare(1,2).
sameSubSquare(2,3).
sameSubSquare(4,5).
sameSubSquare(5,6).
sameSubSquare(7,8).
sameSubSquare(8,9).

%% Make it symmetric and transitive
sameSubSquare(X,Y) :- sameSubSquare(Y,X), position(X), position(Y).\
sameSubSquare(X,Z) :- sameSubSquare(X,Y), sameSubSquare(Y,Z),
position(X), position(Y), position(Z).\

%% Now say there may not be two of the same number in the same subsquare
%% Inequalities not actually needed.
:- state(XA,YA,N), state(XB,YB,N), sameSubSquare(XA,XB),
sameSubSquare(YA,YB), position(XA),
position(XB), position(YA), position(YB),
value(N), XA != XB, YA != YB.

A typical academic year consists of three parts: first semester (Oct-Jan), second semester (Feb-May) and the summer months. For me the first semester is my teaching semester. During these months my main activity is teaching, preparing for teaching and meeting my students. These will also be the months that write up the work that I have been doing during the summer months. My second semester is mainly research although I will be helping students with their projects. The summer months is mostly research. On a research day, I will often talk (in person, via skype or google chat) to other scientists about new ideas or work with them to solve one of our current problems. Alternatively I am writing up results so I can show them to other people working on similar problems. Scientists in a area often meet at conference. These are events that allow scientists to tell colleagues what problems they have solved. I attend normally about 2 of those events a year.  When I am not trying to solve scientific problems, I help my department with administration tasks. I am involved in what we call outreach activities. I go to schools or organise school visits to campus and offer pupils a taste of computer science and/or the life at university.

Occasionally I have the opportunity to visit one of my colleagues abroad and experience what it is like to work in a different university or research lab in a different country.

This is how my diary looked liked when I wrote this text:

9:00: Research meeting with Phd student and colleague in Abbu Dabi
10:00: Skype meeting with colleagues about a research paper that is due in March
11:00: Recruitment and Outreach Meeting
12:00: Working on research paper that is due today
13:00: Lunch
14:00: Meeting with Phd student
14:30: Meeting with Phd student
15:00: Finishing research paper and give it to co-author for checking
16:30: Meeting Phd Student

I normally do not start at 9:00. I tend to be more of an evening person so if I have no meetings I start later and work later in the evening