Presentation 2013-10-19
Exploring Linguistic Features for the Automated Assessment of L2 Spoken English
Yuichiro KOBAYASHI, Mariko ABE,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) The present study aims to automatically evaluate second language (L2) spoken English, and to identify linguistic features for predicting speaking proficiency. It drew on the NICT JLE Corpus, a corpus of 1,281 Japanese EFL learners, coded with speaking proficiency. The oral proficiency levels were used as criterion variable and linguistic features analyzed in Biber (1988) as explanatory variables. Random forests (Breiman, 2001) was employed to predict the speaking proficiency. As a result of random forests with the out-of-bag error estimate, 61.28% of L2 spoken productions were correctly classified. The strongest predictors of an individual's level were tokens, types, prepositions, tense, and first person pronouns.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) L2 Spoken English / Learner Corpus / Machine Learning
Paper # TL2013-41
Date of Issue

Conference Information
Committee TL
Conference Date 2013/10/12(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Vice Chair

Paper Information
Registration To Thought and Language (TL)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Exploring Linguistic Features for the Automated Assessment of L2 Spoken English
Sub Title (in English)
Keyword(1) L2 Spoken English
Keyword(2) Learner Corpus
Keyword(3) Machine Learning
1st Author's Name Yuichiro KOBAYASHI
1st Author's Affiliation Japan Society for the Promotion of Science()
2nd Author's Name Mariko ABE
2nd Author's Affiliation Faculty of Science and Engineering, Chuo University
Date 2013-10-19
Paper # TL2013-41
Volume (vol) vol.113
Number (no) 253
Page pp.pp.-
#Pages 6
Date of Issue