跳到主要內容

臺灣博碩士論文加值系統

(3.236.84.188) 您好!臺灣時間:2021/07/30 03:28
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果 :::

詳目顯示

: 
twitterline
研究生:郭文卿
研究生(外文):Wen-ching Kuo
論文名稱:台語羅馬字聽寫測驗的電腦輔助出題系統研究--以教育部委託辦理台語認證試題研發計畫做研究範圍
論文名稱(外文):A Survey of the Computer-Assisted Taiwanese Dictation Test: A Case Study of the Taiwanese Proficiency Test Project Commissioned by the Ministry of Education
指導教授:蔣為文蔣為文引用關係
指導教授(外文):Wi-vun Chiung
學位類別:碩士
校院名稱:國立成功大學
系所名稱:台灣文學研究所
學門:人文學門
學類:台灣語文學類
論文種類:學術論文
論文出版年:2009
畢業學年度:97
語文別:中文
論文頁數:253
中文關鍵詞:語詞聽寫測驗電腦自動出題音素權重語音特色台語認證
外文關鍵詞:Taiwanese Proficiency TestVocabularyDictation TestAutomatic TestingPhonemeWeight
相關次數:
  • 被引用被引用:4
  • 點閱點閱:353
  • 評分評分:
  • 下載下載:53
  • 收藏至我的研究室書目清單書目收藏:0
現在的社會真注重專業證照,各行各業攏有相關的檢定考試,語言能力測驗嘛是仝款,國內外攏有針對語言類辦理證照考試。尤其國內原住民語、客語、華語攏有專門機構teh辦理認證,唯獨大多數人teh使用的台語無受重視,經過濟濟專家學者的呼籲kap拍拚,教育部佇2008年4月委託成功大學台語認證試題研發團隊,辦理「臺灣閩南語語言能力認證」試題的研發。辦理第1擺全國性的台語認證,團隊的擔頭真重,為著欲減輕試務負擔,針對台語羅馬字語詞聽寫測驗研發自動出題系統kap考生成績的電腦評分作業。系統真精密控制每一個音素的出題次數kap語詞等級,達到電腦化代替人工作業。最主要是確保資訊的正確性,避免錯誤kap疏忽,幫贊團隊的規劃kap執行,順利完成認證任務。
本文是以台語羅馬字語詞聽寫測驗預試為研究範圍,分析資料包括30人樣本考生kap 2擺各500人試考的試題,有效的樣本數有835份。研發團隊所規畫的語言能力分級是參考歐洲理事會CEF的6級標準,測驗方式是採用標準參照測驗,分6級毋閣各級做伙考的原則。團隊安排樣本考生2008年7月27號佇台南試考、2008年8月23 號kap 2008年11月29號分北、中、南、東4區仝時間舉辦各500人的試考。考生的語詞聽寫成績攏透過Excel輸入試考資料結合本系統的運作,真緊就完成考生成績的分析作業。全部的語詞聽寫試考成績是有達到標準,超過80%以上的答對率有469人,佔總數835人的56.17%,證明只要認真學習羅馬字,佇語詞聽寫這個單元是會當得著高分。因為是6級做伙考,測驗題目需要按各級分配語詞出題,台語的詞彙代先有參考華語8,000詞kap詞頻高低制定語詞等級互團隊參考引用,經過第1擺500人的試考分析,發現所議定的語詞等級難易度無合常理;所以本文研擬以台語的語音特色,參考這3擺試考各音素kap複母音的答對率做憑據,透過電腦算出雙音節的語詞參考權重,簡單有效率定義各語詞的等級。
一個好的檢定試驗有足量的詞彙做出題的指引是天經地義的代誌,只是台語的認證先天不足後天失調。本文有收集濟濟詞彙來做資料庫語詞來源,最後是以教育部國語推行委員會於2008年10月公佈的《臺灣閩南語常用詞辭典》網路版為準,經過篩揀全部有9,000詞的雙音節資料,做為語詞聽寫電腦自動出題的語料庫。其中語詞等級的規範是透過統計軟體Spss的K-means公式,將資料分成6組,有效分配各語詞的等級指標,chiah閣修改資料庫的資料。最後得著雙音節的語詞A1級有368筆(權重4.73到6.84);A2級有1,373筆(權重6.85到7.78);B1級有2,674筆(權重7.79到8.64);B2級有2,678筆(權重8.65到9.47);C1級有1,443筆(權重9.48到10.37);C2級有464筆(權重10.38到12.65)。以權重原則所定義的語詞等級,經過系統重新做分析,發覺這個方法會當分出語詞等級的難易度,值得咱進一步研究開發。
針對詞彙等級的定義有需要結合專家學者、社會賢達、使用者共同拍拚,thang有一套逐家攏認同的分級標準kap適當的詞彙量逢引用,按呢chiah會當互語詞聽寫測驗有客觀性kap公信力,達到有水準的語言認證測驗。
Certificates accredited for trades and professions are taken very seriously as a prerequisite of national competitiveness (e.g., employment prospects). Different qualification examinations are specifically designed for every walk of life and language proficiency tests stand no exception. Both public and private organizations administer a variety of language proficiency tests internationally. In Taiwan, different organizations manage the language proficiency tests for specific aboriginal language, Hakka as well as Mandarin Chinese. However, the language test for Taiwanese which is widely spoken for the majority in Taiwan draws less attention.
Acknowledging the importance of standardizing the Taiwanese proficiency of the citizens across the nation, a body of scholars and linguistic experts undertook significant efforts to introduce a test system designed for assessing Taiwanese proficiency. In April of 2008, one well-rounded research team from National Cheng Kung University contracted by Ministry of Education (Henceforth MOE) designed the examination entitled “General Taiwanese Proficiency Test” which determines various levels of competency in listening, writing, reading and speaking.
In order to facilitate the test result electronically, the research team from National Cheng Kung University designed a PC-assisted dictation test system to determine the candidates’ language proficiency of Taiwanese Romanization. This PC-assisted dictation test system is labor-efficient and meticulously designed to test the candidates for each phoneme and varying levels of the difficulty of Taiwanese Romanization accordingly. The objective was to facilitate the examination in terms of result accuracy, administration efficiency as well as accreditation credibility.
This study examined the content of examination paper which was used for the preliminary test of General Taiwanese Proficiency Test. The target participant group comprised 30 candidates who participated in the preliminary test twice successively. Therefore, two corpora of data collected in this study included the participants’ test results for two examinations respectively. This study gathered 835 samples valid for data analysis.
In light of the Common European Framework of Reference for Languages: Learning, Teaching, Assessment (abbreviated as CEFR), the research team devised the guidelines addressing six levels of varying Taiwanese proficiency. The target participant group was asked to sit for the preliminary exam held on the 27th of July in 2008 in Tainan City and later arranged to sit for another preliminary exam held on the 23rd of August and the 29th of November in 2008 in four different examination sites located in the north, the central, the southern and eastern cities across Taiwan. The target participants were dispersed amongst other candidates in the examination site which accommodated 500 examinees at a time in total.
To facilitate the test result, the PC-software Excel was employed to electronically collect each candidate’s test result with the assistance of the test system designed by the research team from National Cheng Kung University. The findings of data analysis showed that all the candidates have met the pass criterion in terms of Dictation Testing wherein 469 examinees held more than 80% of accuracy rate, accounting for 56.17% of 835 examinees. In this respect, the researcher of this study argued that the competence of Taiwanese Romanization plays a crucial role in better performance on Dictation Testing. The difficulty of each test question is designed in accordance with six levels of Taiwanese language proficiency. To determine the varying levels of difficulty of each Taiwanese character, the research team referred to the 8000 Mandarin Chinese characters along with the consideration to how frequently each vocabulary is occurred and spoken on a daily basis.
The analysis of the data collected from the first preliminary test showed that the criterion based on the difficulty of each vocabulary was not able to apply to this study in a fruitful manner. Hence, this study considered the significant speech sounds of Taiwanese and then focused on the examinee’s accuracy of test questions germane to phonemes. With the assistance of SPSS software, this study acknowledged the weight of certain Taiwanese phonemes is associated with varying levels of difficulty of vocabularies.
To successfully assess one’s language proficiency, one test system must refer to the substantial amount of vocabularies while designing the test questions. However, the language testing system for Taiwanese is still on the fledging stage to date. In this respect, this study considered the volume of Taiwanese lexicons from a number of dictionaries and then employed the Contemporary Taiwanese Dictionary (on-line edition) published by National Languages Committee of MOE in October of 2008. A lexicon repertoire of 9000 vocabularies with compound syllables was utilized for the Automatic Testing system.
To facilitate the data analysis, the analytic tool, K-means, within SPSS was statistically applied to divide the 9000 vocabularies into six groups in terms of varying levels of difficulty of vocabulary. The findings showed that 368 vocabularies were categorized as A1 (weight 4.73 to 6.84); 1373 vocabularies were categorized as A2 (weight 6.85 to 7.78); 2674 vocabularies were classified as B1 (weight 7.79 to 8.64); 2678 vocabularies were classified as B2 (weight 8.65 to 9.47); 1443 vocabularies were grouped under C1 (weight 9.48 to 10.37) and 464 vocabularies were grouped under C2 (weight 10.38 to 12.65). The findings of this study rendered a significant insight into the applicability of weight factor while assessing the varying levels of difficulty of vocabularies.
Given the growing awareness of the important role that General Taiwanese Proficiency Test plays in language testing systems, this study acknowledged that the focus on the design of the examination has become critical and relevant. As there are a variety of approaches used to assess the varying levels of difficulty of Taiwanese vocabularies, the researcher of this study believes a credible and just test relies on the invaluable advices and efforts made by the scholars, linguistic experts as well as the language users.
第一章 前言 1
1.1. 研究動機 1
1.2. 研究目的 2
1.3. 研究範圍 3
1.4. 研究方法 4
1.5. 論文架構 5
第二章 文獻回顧 7
2.1. 語言測驗的理論基礎 7
2.1.1. 語言測驗的發展 7
2.1.2. 常模參照測驗vs 標準參照測驗 11
2.1.3. 難度等級 12
2.1.4. 語言檢定的聽寫測驗紹介 15
2.2. 台語音韻系統紹介 16
2.2.1. 台語的音節結構 16
2.2.2. 台語的聲母 18
2.2.3. 台語的韻母 19
2.2.4. 台語的聲調 23
2.2.5. 小結 25
2.3. 語詞聽寫歷史資料分析 26
2.3.1. 詞彙的界定 26
2.3.2. 語詞來源 27
2.3.3. 歷史資料分析 28
2.4. 電腦輔助系統回顧 30
2.4.1. 全民網路英語能力檢定 (NETPAW) 30
2.4.2. AWETS英文字彙自動出題及測驗系統 31
2.4.3. 適性化線上英語聽寫測驗系統之研究 31
2.4.4. 英文聽力字彙與聽寫練習題輔助出題系統之研究 33
第三章 建立台語詞彙語料庫 34
3.1. 現有語料庫紹介 34
3.2. 詞彙篩揀範圍 36
3.3. 詞彙篩揀過程 41
3.4. 小結 47
第四章 語詞聽寫電腦輔助出題系統 48
4.1. 軟硬體需求 48
4.2. 系統架構kap功能 48
4.2.1. 系統架構 48
4.2.2. 資料庫Table設計 54
4.2.3. 系統功能 54
4.3. 語詞聽寫自動出題處理原則 57
4.3.1. 語詞出題原則 58
4.3.2. 語詞等級設定原則 58
4.3.3. 音素設定原則 59
4.3.4. 電腦系統音素處理作業 59
4.3.5. 出題眉角kap考試方式 61
4.4. 考生基本資料輸入輸出 61
4.4.1. 考生基本資料輸入 61
4.4.2. 考生聽寫資料產出 62
4.4.3. 電腦評分的標準 64
4.5. 試考分析 66
4.5.1. 考生分數分析 66
4.5.2. 考生背景分析 72
4.5.3. 各音素答對率分析 81
4.5.4. 語詞答對率分析 87
4.5.4.1. 語詞等級統計 87
4.5.4.2. 語詞明細資料 89
4.5.5. 聲調答錯率分析 92
4.5.6. 小結 94
4.6. 制定音素權重的語詞等級探討 95
4.6.1. 台語語音特色權值設定 95
4.6.2. 重設資料庫語詞等級 98
4.6.3. 重設語詞等級後的分析 100
4.6.4. 小結 112
第五章 結論 113
參考文獻 117
附錄 121
Daniel, S. E. 1996. Client/Server computing, Journal of Systems Maintance, Vol. 47, No. 3, pp. 58-61
Date, C. , 1975. An Introduction to Database System. Reading, Ma: Addison-wesley
Everest, Gorden C. 1986. Database Management Objectives System Functions & Administration, McGraw Hill International Edition
Fellbaum, Christine. (ed.),1998. WordNet: An Electronic Lexical Database. Cambridge, MA: MIT Press.
Glasser.R. 1963. Instructional technology and the measurement of learning outcomes; some questions. American Psychologist, 18, 519-521
Lile, E. A. 1993. Client/Server architecture: a brief overview, Journal of Systems Maintance, Dec. 1993, pp. 26-29
Martin, J. 1977. Computer Database Organization, 2nd ed. Englewood Cliffs, NJ: Prentice-Hall.
Mittra, S. S. 1995. A road map for migrating legacy systems to client/server, Journal of Software Maintance, Vol. 7, No. 2, pp. 117-130
V. Fromkin, R. Rodman, 2002. An Introduction to Language, Thomson Learning, 7th Edition.
V. I. Levenshtein, 1966. Binary codes capable of correctingdeletions, insertions, and reversals, Cybernetics and Control Theory, Vol. 10, No. 8, pp. 707-710, 1966.
王育德,2000,《臺灣話講座》,台北,前衛出版社。
王育德,2002,《臺灣語研究卷》,台北,前衛出版社。
余民寧,2002,《教育測驗與評量:成就測驗與教學評量》,台北,心理。
吳琮璠、謝清佳,1996,《資訊管理理論與實務》,台北,松崗電腦圖書。
李勤岸,2006,《母語教育政策及拼音規劃》,台南,金安文教。
周文欽、歐滄和、許擇基、盧欽銘、金樹人、范德鑫,1995,《心理與教育測驗》,台北,心理出版社。
邱文錫、陳獻國,2000,《實用華語臺語對照典》台北,樟樹出版社。
柯華葳,2004,《華語文能力測驗編製-研究與實務》,台北,遠流出版社。
莊永山,2007,《歐洲共同語文參考架構 The Commom European Framework of Reference for language: Learning, Teaching, Assessment》,高雄,和遠。
洪惟仁,1985,《臺灣河佬話聲調研究》,台北,自立晚報。
張凱,2002,《語言測驗理論與實踐》,北京,北京語言文化大學出版設社。
張凱,2006,《語言測試理論及漢語測試研究》,北京,商務印書館。
張復聚、王淑珍,2001,《台灣字ABC》,台北,台語傳播企業。
教育部,2007,《臺灣閩南語羅馬字拼音方案使用手冊》,台北,教育部。
許極燉,1990,《台灣語概論》,台北,台灣語文研究發展基金會。
郭生玉,1985,《心理與教育測驗》,台北,精華。
陳修,2000,《台灣話大詞典》,台北,遠流出版社。
董峰政,2000,《台語文天地》,台南,立誠印刷。
榮泰生,1990,《管理資訊系統》,台北,松崗電腦圖書資料公司。
劉潤清、韓寶成編著,1999,《語言測試和它的方法》修訂版,北京:外語教學與研究出版社。
蔣為文,2005,《語言、認同與去殖民》,台南,成大。
蔣為文,2007,《Gí-giân, Bûn-hãk kap Tâi-oân Kok-ka Chài-sióng-siöng╱語言、文學kap台灣國家再想像》,台南,成大。
鄭良偉,1997,《台語、華語的結構及動向Ⅰ:台語的語音與詞法》,台北,遠流出版社。
鄭良偉、曾金金、李櫻、盧廣誠,2000,《大學台語文選》,台北,遠流出版社。
王俊弘、劉昭麟、高照明,<電腦輔助英文字彙出題系統之研究>,國科會計畫NSC-92-2213-E-004-004。
論文資料
李台元,<台灣原住民族語言能力認証制度之評估>碩士論文,2003。
李國賢,<英語科新式選擇題型之分析研究>,碩士論文,1991。
林美雪,<第一次台語認證試題研發工作坊論文集>,2008,台南市:台灣國家文學館。
林瑞青,<招募資訊系統之建立與發展—以H相業銀行為例>碩士論文,2001。
張莉萍,<華語詞彙分級初探>,中國廈門「第六屆漢語辭彙語意學研討會」, 2005.04.20-23。
陳盛林,<管理資訊系統評估機制設計之研究>碩士論文, 1995。
黃上銘,<英文聽力字彙與聽寫練習題輔助出題系統之研究>碩士論文, 2006。
黃上銘、劉昭麟、高照明,<適性化線上英語聽寫測驗系統之研究>,國科會計畫NSC-92-2213-E-004-004。
黃居仁,2003,<知識的投射與解讀:兼論數位典藏中之語言座標>,漢學研究通訊。20:2(總第78期)頁83-89。
楊允言,<台語文語料庫蒐集及語料庫為本台語書面語音節詞頻統>,國科會計畫,2004.8 -- 2005.7。
臺北縣政府教育局,<臺北縣國小閩南語1000常用詞彙如何配合九年一貫課程之研究>,2003.12。
台灣閩南語語言能力認證學術研討會論文集,2009,台南,台灣國家文學館。
蔡培火,<我在文化運動所定的目標>,《臺灣民報》,1927.1.2
鄭良偉,<台華對譯詞庫>,http://taigu.eic.nctu.edu.tw/ ,20090408資料
鄭良偉,<台灣語文賞析>,交通大學通識課程的教材
鄭錦全,<從計量理解語言認知>,收錄於Benjomin Tson, Tomlai Sam Chan, William Wang 編《漢語計量與計算研究》,頁15-30,香港城市大學,1998。
簡茂發,<命題方法與試題分析>,國教輔導,第31卷第一期,頁2-13,1991。
連結至畢業學校之論文網頁點我開啟連結
註: 此連結為研究生畢業學校所提供,不一定有電子全文可供下載,若連結有誤,請點選上方之〝勘誤回報〞功能,我們會盡快修正,謝謝!
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top