Logo image
語音辨識技術支援之線上會話教材對大學生英文學習之效能研究
Thesis

語音辨識技術支援之線上會話教材對大學生英文學習之效能研究

邱作麟
Masters, National Tsing Hua University
2005

Abstract

電腦輔助語言教學語音辨識語言行為溝通能力會話 Computer Assisted Lanuage LearningAutomatic Speech RecognitionSpeech ActCommunicative CompetenceConversation
With growing emphases on communicative competence, teaching of speech acts is essential for learners to use language appropriately in various contexts. It has been commonly believed that the competence of using speech acts appropriately is better enhanced when learning in native-speaking contexts and having face-to-face conversations with native speakers. Unfortunately, for most EFL (English as Foreign Language) learners, these conditions are rarely easy to attain. In recent years, one promising field in Computer Assisted Language Learning (CALL) is the application of Automatic Speech Recognition (ASR) technology for assisting learners to engage in meaningful speech interactions. Such technology facilitated the development of simulated real-life conversations, in which the application of ASR offers judgment on student performance. Therefore, a web-based conversation environment, CandleTalk (http://candle.cs.nthu.edu.tw), which allows learners to seemingly talk with the computer, has been developed to help EFL learners receive explicit training on speech acts and thus demonstrate communicative competence. CandleTalk is equipped with an ASR engine that judges whether learners provide appropriate input in the interactions. Six speech acts--greeting, parting, requesting, apologizing, complaining, and complimenting--are presented as the foci of four units. Moreover, local information is integrated in the content of the dialogues to strengthen student motivation. A study was conducted on 49 college freshmen, 29 English majors and 20 non-English majors from National Tsing Hua University, to investigate their learning process in practicing with the units within a period of 6 weeks and explore their perception towards CandleTalk. Two tests, a pretest and a post using the format of spoken Discourse Completion Test (DCT), were administered for evaluating students’ learning outcomes from practicing in the environment. An evaluation questionnaire was used for understanding students’ perception toward three dimensions of CandleTalk: (1) the contents of the four units, (2) functunality, and (3) overall effectiveness of the site. The results of the study show that the application of ASR is effective for college freshmen in the learning of speech acts. From the comparison of pretest and posttest, students overall showed significant improvement in their overall scores and the use of speech acts, with more improvement found in non-English majors. Moreover, learners in general perceived positively towards the instruction supported by ASR, though they also suggested further improvement on the recognition rate and the recording procedure. Finally, pedagogical implications suggest language teachers’ judicious use of ASR technology for providing alternatives for EFL learners’ oral practice. With regard to developing ASR-supported CALL environment, future developers should figure out innovative ways of material presentation for coping with current recognition limit of the technology. The input the environment offers, chances for output production, and pertinent feedback provided are three key elements for developing an effective conversation environment.

Metrics

1 Record Views

Details

Logo image