本文是一篇研究語(yǔ)音識(shí)別技術(shù)輔助漢英交替?zhèn)髯g的實(shí)驗(yàn)報(bào)告。報(bào)告以吉爾的交替?zhèn)髯g精力分配模型為理論基礎(chǔ),探究了語(yǔ)音識(shí)別技術(shù)輔助漢英交替?zhèn)髯g的可能性,分析語(yǔ)音識(shí)別技術(shù)對(duì)人類(lèi)口譯員的影響,并對(duì)該技術(shù)未來(lái)的發(fā)展應(yīng)用提供建議。筆者采用了目前市面上識(shí)別準(zhǔn)確率較高的科大訊飛語(yǔ)音識(shí)別軟件將中文發(fā)言轉(zhuǎn)錄為文本,并利用視頻編輯軟件按照訊飛語(yǔ)音識(shí)別軟件識(shí)別速度將文本制作成字幕供實(shí)驗(yàn)使用。筆者將受試譯員分為兩組,實(shí)驗(yàn)組使用訊飛語(yǔ)音識(shí)別軟件協(xié)助進(jìn)行交傳,對(duì)照組不使用任何輔助工具進(jìn)行交傳。實(shí)驗(yàn)結(jié)束后,筆者收集分析實(shí)驗(yàn)數(shù)據(jù)并對(duì)實(shí)驗(yàn)組受試者進(jìn)行回訪。實(shí)驗(yàn)結(jié)果表明,語(yǔ)音識(shí)別技術(shù)能夠減輕譯員的記憶負(fù)荷,提高譯文的準(zhǔn)確度、流暢度及質(zhì)量;但文本錯(cuò)誤可能對(duì)譯員造成負(fù)面影響,影響譯文質(zhì)量;總的來(lái)說(shuō),人類(lèi)譯員與語(yǔ)音識(shí)別技術(shù)的結(jié)合有助于減輕譯員的記憶負(fù)荷,提高譯文質(zhì)量,但語(yǔ)音識(shí)別軟件的準(zhǔn)確度及識(shí)別錯(cuò)誤的能力有待進(jìn)一步提高。

【文章頁(yè)數(shù)】：89 頁(yè)

【學(xué)位級(jí)別】：碩士

【文章目錄】：
ACKOWLEDGEMENTS
ABSTRACT
摘要
1.INTTRODUCTION
    1.1 Background
    1.2 Layout of the thesis
2.LITERATURE REVIEW
    2.1 Development of machine-aided translation
    2.2 Development of speech recognition
    2.3 Relevant literature on interpreting technologies
3.THEORETICAL FRAMEWORK
    3.1 Gile's effort model
    3.2 Introduction to iFlytek automatic speech recognition
4.EXPERIMENT DESCRIPTION
    4.1 Objectives
    4.2 Methodology
    4.3 Material
    4.4 Participants
    4.5 Procedures
5.EXPERIMENTAL RESULTS AND ANALYSIS
    5.1 Influence on accuracy rate
    5.2 Influence on fluency
    5.3 Influence of incorrect scripts
        5.3.1 Incorrect numbers
        5.3.2 Incorrect terms
    5.4 Coordination
    5.5 Interpreters'feedback
    5.6 Experimental conclusion
        5.6.1 Major findings
        5.6.2 Implications and suggestions
        5.6.3 Limitations of the experiment
6.CONCLUSION
REFERENCES
APPENDICE
    Appendix A
    Appendix B



本文編號(hào)：3922899