Edinburgh Research Explorer

I wanna talk like you: Speaker Adaptation to Dialogue Style in L2 Practice Conversation

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Related Edinburgh Organisations

Open Access permissions



Original languageEnglish
Title of host publicationProceedings of the 20th International Conference on Artificial Intelligence in Education
EditorsSeiji Isotani, Eva Millán, Amy Ogan, Bruce McLaren, Rose Luckin
PublisherSpringer, Cham
ChapterPart II
Number of pages5
ISBN (Electronic)978-3-030-23207-8
ISBN (Print)978-3-030-23206-1
Publication statusPublished - 21 Jun 2019
Event20th International Conference on Artificial Intelligence in Education - Chicago, United States
Duration: 25 Jun 201929 Jun 2019

Publication series

NameLecture Notes in Computer Science
PublisherSpringer, Cham
ISSN (Electronic)0302-9743


Conference20th International Conference on Artificial Intelligence in Education
Abbreviated titleAIED 2019
CountryUnited States
Internet address


We present a novel method for analysing speaker alignment in second language conversational practice dialogue. In particular, we represent utterances as Dialogue Acts, and use Epistemic Network Analysis to analyse their usage patterns. We hypothesise convergence between speakers will be visible at this level, and influenced by learner ability and tutor strategy. We find as student ability increases, the distribution of student and tutor Dialogue Acts is both initially closer, and converges more over the course of the dialogue. We also find that the type of Dialogue Act use changes both with ability and throughout an interaction. Our results contribute a novel method for the automatic analysis of both learner ability and tutor strategy, which can inform both the development of personalised automatic tutoring tools, and formative assessment and feedback in an educational setting.

    Research areas

  • epistemic network analysis, scaffolding, dialogue, natural language processing, alignment, Zone of proximal development


20th International Conference on Artificial Intelligence in Education


Chicago, United States

Event: Conference

Download statistics

No data available

ID: 82481560