國小學童於網路合作學習中對話行為之自動分類系統

陳昭安

指導教授 ： 邱瓊慧

國立臺灣師範大學/教育學院/資訊教育學系/碩士(2010年)

摘要

本研究建構一自動化的對話分類系統，使其能將國小學童所進行之電腦合作學習活動的對話內容，依照已定義的對話行為類型進行自動分類，包括如｢討論概念｣、｢討論命題｣、｢整理階層關係｣、｢要求幫助或提供幫助｣、｢檢討構圖任務的正確性｣、｢時間控制｣、｢流程控制｣、｢協調操作權責｣、｢計較權責｣、｢激勵組員｣、｢確認組員是否在線上｣、｢詢問組員身分｣、｢搗亂｣、｢吵嘴｣、｢其它事件｣等十五類的對話行為類型。此系統先將對話內容進行斷詞，經由本體論中相關權重運算之後，能將對話內容歸類到各所屬的對話行為類型。為了測試該系統的分類效果，將電腦所分類出的結果與人工分析結果進行比對。研究結果發現：（a）當該系統在分類｢檢討構圖任務的正確性｣及｢流程控制｣的對話行為類型時的結果較佳，但在分類｢詢問組員身分｣、｢其它事件｣的對話行為類型時的結果較差；（b）系統針對無完整結構的句子、需從溝通脈絡判斷對話行為類型的句子以及有錯別字的句子容易分類錯誤，建議未來設計應能考量加入錯別字的校正以及對話內容脈絡的處理。

關鍵字

電腦合作學習 ； 本體論 ； 對話行為

並列摘要

An automatic classification system for primary school students' dialogues was developed in the study. The system categorized participants’ dialogues obtained from a study on computer supported collaborative learning based on 15 types of dialogue acts, including “concept”, “proposition”, “hierarchy”, “for help and help”, “review”, “time controlling”, “flow controlling”, “conflict dealing”, “accountability demanding”, “motivating”, “name calling”, “disturbing”, “bicker” and “others”, by word segmentation, weighting calculation and classification from the perspective of ontology. The classification produced by the system was compared with the classification made by the experts in order to examine the effectiveness of the system. The study result revealed that (a) dialogues related to the dialogue act “review” and “flow controlling” could be classified more properly by the system than dialogues belonged to the dialogue act “name calling” and “others”; and (b) sentences, without structures or with misspelled words, and dialog contained context tended to be classified improperly by the system. Therefore, an advanced system with grammatical and spelling check and detection of context was recommended for the future development.

並列關鍵字

computer supported collaborative learning ； ontology ； dialogue act

參考文獻

Barros, B., & Verdejo, M. F. (2000). Analysing student interaction processes in order to improve collaboration. International Journal of Artificial Intelligence in Education, 11, 221-241.

Avancini, H., Lavelli, A., Sebastiani, F., & Zanoli, R. (2006). Automatic expansion of domain-specific lexicons by term categorization. ACM Trans. Speech Lang. Process., 3(1), 1-30.

Chiu, C.-H. (2003). Exploring how primary school students function in computer supported collaborative learning. Engineering Education and Lifelong Learning, 13(3/4), 258-267.

Fišel, M. (2007). Machine learning techniques in dialogue act recognition. Estonian Papers in Applied Linguistics(3/2007), 117-134.

Goodman, B. A., Linton, F. N., Gaimari, R. D., Hitzeman, J. M., Ross, H. J., & Zarrella, G. (2006). Using dialogue features to predict trouble during collaborative learning. User Modeling and User-Adapted Interaction, 16(1), 83-84.