"European Accent Map"

Supervised by M.Cooke

Abstract

This project aims to build a resource of speech information that relates specifically to the different accents found in regional speaking. The resource will aid research, teaching and help build the body of speech knowledge. The purpose of this project is to develop a software system to enable the capture of speech samples and personal background from a geographically, culturally and socially diverse group of people, and to build a usable resource of that information.

An object-oriented approach was taken in the project, supported by techniques such as UML, Java, JavaSound, XML, SQL and traditional testing methods. Importance was placed on the usability; a critical aspect for what will be a publicly accessible system. This project provides a fully working system allowing the capture, searching and presentation of speech samples and background information. The system is designed for general use by many linguists, who each can define their own lists and provide a fully localised interface. The foundations are laid for further work to manipulate the resource.