翻譯研究所學生使用機器翻譯之意圖與接受度初探—以全臺翻譯研究所學生為例

Abstract

近年來機器翻譯與類神經技術的結合與發展，在翻譯產業掀起了一陣波瀾，改變了翻譯產業的生態及譯者工作的模式。鑑於機器翻譯與譯者工作的連結愈來愈緊密，產業的相關需求也不斷提升（Slator, 2021; DePalma et al., 2021），許多翻譯學者（Mellinger, 2017）紛紛呼籲，翻譯教育應納入翻譯科技能力的相關訓練，以確保學生在未來自動化科技發展的浪潮下，仍能維持專業譯者的市場競爭力。本研究以全臺九所授予翻譯碩士學位學校之學生為研究對象，以Davis（1989）提出之科技接受度模型（Technology Acceptance Model）為基礎，結合過去相關研究實證之外部變項，採問卷調查法並結合半結構式訪談，初探目前翻譯人才的機器翻譯使用與接受度現況，並試圖探討影響學生機器翻譯使用與接受度的關鍵因素，分析當前各大翻譯學校（碩士學位層級）提供的訓練如何影響學生的機器翻譯使用與態度。

本研究問卷於2021年10月至11月進行發放，共回收79份有效問卷，並自問卷受訪者中選擇10位進行半結構式訪談。資料經統計分析所得結果如下：

（一） 高達98.73%的受訪學生具有機器翻譯使用經驗。 （二） 受訪學生的機器翻譯使用意圖（即接受度）非常高。 （三） 知覺有用為影響受訪學生機器翻譯接受度的關鍵因素，而知覺有用又顯著受知覺易用、工作關聯等外部變項影響。 （四） 信任及對機器翻譯的恐懼會接影響受訪學生的機器翻譯使用意圖。 （五） 機器翻譯相關教課程對受訪學生的機器翻譯接受度無正向影響，惟具備電腦輔助翻譯工具訓練經驗者，機器翻譯接受度則顯著高於其餘受訪者。研究結果顯示，全臺翻譯相關系所（碩士層級）學生的機器翻譯使用頻率與接受度皆非常高，而科技接受度模型也驗證了許多影響其接受度高低的關鍵因素，針對目前全臺翻譯相關系所（碩士層級）開設之機器翻譯相關課程所面臨的侷限，提供了實質建議與未來發展方向。In recent years, machine translation's (MT) integration with artificial neural networks has sparked discussions within the translation industry. The unprecedented advancement in MT technology has transformed the industry dynamics as well as how translators work.

In view of MT's increasing impact on translation practices and the industry's increasing demands in MT-related services (Slator, 2021; DePalma et al., 2021), translation scholars (Mellinger, 2017) have called on translation institutes to formalize machine translationtraining in an effort to ensure professional translators' competitiveness in a future increasingly dominated by automation. This study aims to explore the use and acceptance of MT, as well as the key factors behind the attitudes and intention to use among translation students enrolled in the Translation and Interpretation (T&I) master's programs across Taiwan through an extended Technology Acceptance Model (TAM) and semi-structured interviews. The model used in this study is based on Davis's (1989) Technology Acceptance Model and is incorporated with potential factors that were found significant in previous studies. Through locating the key factors, this study seeks to understand how the trainings currently offered at the T&I institutes affects translation students' attitude and intention to use MT in Taiwan.

A total of 79 surveys were collected from 9 T&I schools in Taiwan between October to November, 2021; and 10 survey respondents were selected for semi-structured interviews. The key findings were as follows:

1. 98.73% of the translation students surveyed have experience using MT. 2. The translation students' intention to use (or acceptance of) MT is rather strong. 3. Perceived usefulness is the key factor behind translation students' intention to use MT; and translation students' perceived usefulness is significantly affected by perceived ease of use, job relevance and other external factors. 4. Trust of quality and fear of MT's growing influence directly affects translation students' intention to use MT. 5. MT-related trainings offered at the T&I schools do not have a positive effect on translation students' intention to use MT. Though, translation students with Computer-Aided Translation Tools (CAT Tools) training scored significantly higher in their intention to use MT than those without any CTA Tools training.The results showed that the frequency and acceptance of MT among translation students in Taiwan are rather high. The proposed TAM model successfully validated a number of key factors behind their acceptance of MT. In response to the current constraints of the MT-related trainings in Taiwan, the results indeed shed light on the possible enhancements of the future trainings at the T&I institutes.