Automated Assessment of Speech Fluency for L2 English Learners

Autor/in	Yoon, Su-Youn
Titel	Automated Assessment of Speech Fluency for L2 English Learners
Quelle	(2009), (138 Seiten) PDF als Volltext Verfügbarkeit Ph.D. Dissertation, University of Illinois at Urbana-Champaign
Sprache	englisch
Dokumenttyp	gedruckt; online; Monographie
ISBN	978-1-1095-8140-9
Schlagwörter	Hochschulschrift; Dissertation; Phonemes; Second Language Learning; Scoring; Correlation; Language Fluency; Computer Assisted Testing; Speech; Student Evaluation; English (Second Language); Pronunciation; Grammar; Phrase Structure; Error Patterns; Regression (Statistics); Models; Computer Software; Educational Technology + Suchen Sie Ihr Suchwort? Thesis; Dissertations; Academic thesis; Fonem; Zweitsprachenerwerb; Bewertung; Korrelation; Language skill; Language skills; Sprachkompetenz; Speaking; Sprechen; Schulnote; Studentische Bewertung; English as second language; English; Second Language; Englisch als Zweitsprache; Aussprache; Grammatik; Phrasenstruktur; Fehlertyp; Regression; Regressionsanalyse; Analogiemodell; Unterrichtsmedien
Abstract	This dissertation provides an automated scoring method of speech fluency for second language learners of English (L2 learners) based that uses speech recognition technology. Non-standard pronunciation, frequent disfluencies, faulty grammar, and inappropriate lexical choices are crucial characteristics of L2 learners' speech. Due to the ease of automatic feature extraction, this study focused on quantitative measures for disfluencies and pronunciation errors. This study developed automated methods for temporal features and pronunciation error detection. In addition to temporal features, clause-internal (non-juncture) disfluency features were calculated automatically. Previous studies found a strong negative correlation between non-juncture disfluencies and fluent speech. In order to implement this relationship in the automated scoring, an automated non-juncture pause detection method was developed, and the quantitative features based on the frequency and duration of the non-juncture disfluencies were calculated. A multi-regression model was constructed using automatically extracted temporal features. The rate of speech was the best predictor, and there was a moderate improvement by adding non-juncture features to the model. The results suggested that non-juncture features can improve the predictability of the automated scoring method. An automated pronunciation scoring system was developed based on confidence scoring method and the classifier. The phonemes where L2 English learners make frequent pronunciation errors were selected, and SVMs were trained in order to distinguish the correct phonemes from their frequent substitution errors. Using landmark-based SVMs, the method was specialized for phonemes where L2 English learners make frequent errors. The automated scoring method in this study will be a useful method not only for automated scoring but also for interactive training. For example, the automated pronunciation scoring method can be used independently as an interactive pronunciation training tool. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com/en-US/products/dissertations/individuals.shtml.] (As Provided).
Anmerkungen	ProQuest LLC. 789 East Eisenhower Parkway, P.O. Box 1346, Ann Arbor, MI 48106. Tel: 800-521-0600; Web site: http://www.proquest.com/en-US/products/dissertations/individuals.shtml
Erfasst von	ERIC (Education Resources Information Center), Washington, DC
Update	2017/4/10