Fluent

Fluent Research

Foreign Language Utilities for Enhanced Learning and Teaching (FLUENT) is a project which aims to revolutionise learning to speak a second language. Focusing on real speech patterns, FLUENT is intended to complement and integrate with, rather than replace, existing language learning courses and materials. The research centres on new methodologies, software and materials to help students and teachers of foreign languages develop listening and speaking skills when dealing with native-to-native (or fluent non-native to non-native) speech.

Project Details

PRINCIPAL INVESTIGATOR

 

 

START DATE

March 2008

COLLABORATORS

END DATE

Ongoing

PROJECT PARTNERS

Language Group

FUNDING SOURCE

Enterprise Ireland

Technical Information

FLUENT’s methodology and software can be used for any language but the archetype used is English. FLUENT addresses the many issues that users of current learning materials experience. Existing courses and materials tend not to consider matters such as elisions (e.g. “gonna” instead of “going to”), regional accents, pragmatics and the dynamic between speakers. FLUENT’s primary mechanism for addressing this is research into, and development of, a Dynamic Speech Corpus (DSC) for authentic Spoken English. Allied to this is a methodology for dealing with natural speech as well as innovative software tools with which the user can access the corpus.

A suitable tagging framework, specifically designed for speech and its associated paralinguistic features, has been researched and built. A software toolkit has been developed to allow a linguist to transcribe the orthotext of the conversation, mark interesting features in speech patterns and to have this immediately associated with the speech and entered into the corpus. This software is being integrated into a seamless platform for technology-enhanced language learning. Students can listen to, interact with, or pause a conversation to find relevant linguistic information about what they heard. A series of exemplar lessons are in development to demonstrate how best to use the system.

Achievements

Initial linguistic specification for the Dynamic Speech Corpus (DSC) completed (March 2008)

Initial technical specification for the prototype DSC completed (June 2008)

Speech Slow-Down technology substantially enhanced (September 2008)

Implementation of the initial, simplified linguistic structure in XML completed (November 2008)

Working prototype showing the feasibility of the approach successfully completed (December 2008)

Detailed consideration given to pedagogical issues in the use of the system (February 2009)

Novel structured approach researched & developed to lead students from existing commercial materials to full-speed conversations between native-to-native (or fluent non-native to non-native) speakers (April 2009)

Extremely detailed hierarchical linguistic structure, based on analysis of a series of conversations, researched and completed (June 2009)

Implementation of full linguistic structure in XML completed (September 2009)

Detailed software Toolset, for use in transcribing and mark-up of conversations, researched, developed & deployed (November 2009)

Working version of user interface, allowing student and teacher access to the corpus and to associated learning materials, is currently being completed

Initial set of exemplar exercises has been developed and is currently being expanded

Conference Papers

Campbell, D. Meinardi, M. Richardson, B. Wang, Y. & McDonnell, C., 2009. DIT’s Dynamic Speech Corpus and dialogic fluency. In: EUROCALL (European Association for Computer-Assisted Language Learning), EuroCALL 2009: new trends in CALL: working together. Gandia, Spain 9-12 September 2009.

Campbell, D. Meinardi, M. Richardson, B. & Wang, Y., 2009. DIT’s Dynamic Speech Corpus and the new learning paradigm. In: IATEFL (International Association of Teachers of English as a Foreign Language), 43rd IATEFL Conference. Cardiff, UK 1-4 April 2009.

Campbell, D. Wang, Y. McDonnell, C, Meinardi, M. & Richardson, B., 2009. The DIT Dynamic Speech Corpus. In: IVACS (Inter-Varietal Applied Corpus Studies), IVACS Symposium Conference: corpus and interaction. Edinburgh, UK 22nd January 2009.

Campbell, D. Wang, Y. & McDonnell, C., 2008. FS ≠ FS: formulaicity and prosody. In: BAAL (British Association for Applied Linguistics), 41st BAAL annual meeting. Swansea, UK 11-13 September 2008.

Campbell, D. Wang, Y. & McDonnell, C., 2007. A prototype speech corpus. In: EUROCALL (European Association for Computer-Assisted Language Learning), EuroCALL 2007: mastering multimedia: teaching languages through technology. Coleraine, UK 5-8 September 2007.

Campbell, D. McDonnell, C. Wang, Y. Meinardi, M. Richardson, B. & Pritchard, C., 2007. DITCall-Slow: slowing native speech for language learners. In: ILTA (the Irish Learning Technology Association), EdTech2007: 8th annual Irish technology users conference. Dublin, Ireland 23-25 May 2007.

Publications

Campbell, D. McDonnell, C. Meinardi, M. Pritchard, C. Richardson, B. & Wang, Y., 2009. DIT’s Dynamic Speech Corpus. Speak Out!, 41, pp.8-11.

Campbell, D. McDonnell, C. Meinardi, M. Pritchard, C. Richardson, B. & Wang, Y., 2009. DIT’s Dynamic Speech Corpus. Proceedings of the 43rd IATEFL 2009 Conference.

Campbell, D. Wang, Y. & McDonnell, C., 2008. FS ≠ FS (Formulaicity and Prosody). Proceedings of the 41th BAAL 2008 Conference.

Campbell, D. McDonnell, C. Wang, Y. Meinardi, M. Richardson, B. & Pritchard, C., 2007. DITCall-Slow: slowing native speech for language learners. Proceedings of the EdTech 2007 Conference.

Digital Media Centre
DIT Aungier Street
Dublin 2
Ireland

tel +353 (0)1 402 3092
fax +353 (0)1 402 3293
info@dmc.dit.ie

view map
dmc

© DMC 2010