Student project: Prolog representation for music

Suitable For

M.Sc. Conversion (Research Project)
M.Sc. Global
4th year final project
3d year final project

Number of People

Synopsis

Written music has a number of characteristics which make it an ideal testbed for data mining techniques in dynamic domains. There is a wide range of levels of abstraction: from the notes in the score (or simple time-stamped events as in MIDI-files), via motifs, phrases, themes, chords, cadenses, up to the formal level of musical structure. Learning in such domains could be seen as raising the abstraction level of the input data.

It is well-known that the success or failure of learning techniques is tightly connected to the representation chosen. This project aims at proposing a Prolog representation for music that does justice to all the possible levels of abstraction, and provides programs for transforming between these levels, including the lowest level of MIDI files.

Once the representation has been chosen and implemented, the aim will be to apply existing inductive logic programming techniques in order to learn complex musical structures from data.

Objectives

To explore the problem of music representation from an AI perspective and propose and implement a representation in Prolog.

Deliverables

An operational environment written in Prolog for reading and writing MIDI files and transform the data between various levels of abstraction.

Skills Required

competence in music (MIDI, notation, a bit of analysis)
competence in Prolog and a general understanding of knowledge representation

Tools/Environments

Prolog (platform still to be decided)

Benefit to Student

To take part in some genuine and valuable artificial intelligence research
Develop rapid prototyping skills
Being able to employ musical skills in a computer science project

Contact

Peter.Flach@cs.bris.ac.uk

Background

http://www.cs.bris.ac.uk/Research/MachineLearning/