透過您的圖書館登入
IP:18.117.107.90
  • 學位論文

利用知識地圖與資訊量計算的基本發問策略設計-以行動學習的導覽代理人的應用為例

Utilize Knowledge Map and Entropy in the Design of Basic Questioning -- An Application to the Guiding Agent of Mobile Learning

指導教授 : 賀嘉生
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


在行動學習中是學生實際到一個地方,針對學習的目標在此地進行學習、觀察,利用行動設備在網際網路(Internet)上進行資料的紀錄以及查詢。利用這樣的機制,在本論文中希望設計一個發問以及導引的機制。發問的機制能使用合成語句非題庫系統的方式進行發問,依照學生答題的狀況,再引導學生到此地,進行學習或觀察何種概念。 本論文中是利用知識地圖(Knowledge Map)儲存知識的架構,分析教材中的關鍵字儲存在知識地圖的架構中,利用這樣儲存知識的架構,找出知識地圖有哪幾種的發問方式以及如何去計算各個概念的資訊量(Entropy)。為了避免學生在同一次導覽中一直接收到同樣的概念,所以在計算各概念的資訊量時,會依據學生的答題狀況根據邊際效益遞減律(Law of Diminishing Marginal Utility)的想法,在每次發問時對概念資訊量進行遞減。在題目中會包含所要測驗的概念,知曉題目中所測驗的概念後,搭配概念的資訊量可進而求出一個題目的資訊量。計算所出題目每題的資訊量後再選出最大資訊量的題目進行出題,並交由智慧型代理人(Intelligent Agent)出題以引導學生進行學習。在學生答完題後,會依據答題狀況再去導引學生到下一個地點進行學習。 在智慧型代理人中,知識地圖出題是利用知識指令(Knowledge Instructions)將概念與知識指令結合,而知識指令會在知識地圖中操作,於知識地圖中取得該題目所含的概念。根據轉換生成語法(Transformational - Generative Grammar),將知識指令合成語句,再把這樣一個語句轉化為問句的方式。 本論文實作了一個發問以及導引系統,搭配桃園縣成功國小校園植物花圃的資料,進行處理。將此發問以及導引系統套用到一個行動學習平台上,搭配實際教案進行實地測試,針對此次的實地測試也整理出此系統有什麼樣需改進的問題。

並列摘要


In Mobile Learning, students learn and observe objects for the learning goal in the place, and use mobile equipments to inquire and to record data on Internet. By this mechanism, this paper designs a system to ask questions and to guide students where to learn. The mechanism of question can compose sentence, not question base, to ask student. According to a status of the student’s answers, the system guides the student to this position to learn and to observe what kinds of concept. This paper uses the structure of Knowledge Map to store knowledge. We analyse keywords in the context and store them into Knowledge Map structure. By this knowledge structure, we can find out what kinds of questioning styles and how to compute the Entropy of each concept. According to the idea of the law of diminishing marginal utility, we stand on the status of the students’ answers to decrease the Entropy of concept gradually. It avoided students receiving the same concepts in the guidance. It will include testing concepts in the question so that we can compute the question’s Entropy by concept’s Entropy after we knowing the testing concepts. The Intelligent Agent will choose the maximum Entropy of question to guide students to learn. After answering questions, the system will estimate the students’ answers to guide them to next position. The Intelligent Agent produces questions by knowledge instructions. The knowledge instructions, which operates in Knowledge Map, composes concepts which has gotten from Knowledge Map. Then according to Transformational-Generative Grammar, it composes knowledge instructions into a sentence and transforms the sentence into a question. This paper implements a system for question and guidance and handles the plants’ data of Cherng Gong Elementary School in Taoyuan County. We integrated this system into a Mobile Learning system (Knowledge-Enhanced System, KEN), and went to Cherng Gong Elementary School to test this system with a lesson plan. We marshaled some problems in this test.

參考文獻


[林伯成2002] 林伯成(2002)。利用知識地圖診斷數學問題之研究。未出版之碩士論文,私立中原大學資訊工程學系,桃園縣,台灣,中華民國。
[郭立平2002] 郭立平(2002)。利用知識地圖與問題解決方法設計出題系統。未出版之碩士論文,私立中原大學資訊工程學系,桃園縣,台灣,中華民國。
[陳惠茹2004] 陳惠茹(2004)。設計對話型代理人語言協助引導認知學習。未出版之碩士論文,私立中原大學資訊工程學系,桃園縣,台灣,中華民國。
[董大賢2002] 董大賢(2002)。利用知識地圖發展物理解題系統之研究。未出版之碩士論文,私立中原大學資訊工程學系,桃園縣,台灣,中華民國。
[Aus1968] David P. Ausubel (1968). Educational Psychology: A Cognitive View. New York: Holt, Rinehart & Winston.

延伸閱讀