少即是多：中小學科學教科書關鍵詞彙探勘之研究（S901）Less is more: A study on retrieving keywords in the primary and middle school science textbooks
| 計畫屬性 | 院外研究計畫 |
|---|---|
| 計畫編號 | MOST 104-2511-S-656-005- |
| 計畫名稱 | 少即是多：中小學科學教科書關鍵詞彙探勘之研究（S901） |
| 計畫名稱(外文) | Less is more: A study on retrieving keywords in the primary and middle school science textbooks |
| 計畫類型 | 個別型計畫 |
| 計畫主持人 | 彭致翎 Peng, Chih-Ling |
| 共同主持人 | 吳鑑城 Wu, Jian-Cheng |
| 經費來源 | 科技部 |
| 執行方式 | 學術補助(國科會專題研究補助) |
| 年度 | 2015 |
| 執行期程(起) | 2015-11-01 |
| 執行期程(訖) | 2016-12-31 |
| 執行狀態 | 已結案 |
| 計畫經費 | 467000 |
| 摘要 | 在科技日新月異、資訊爆炸時代，如何在龐雜巨量資訊中，有效擷取重要知識，進行有意義的理解學習，是現代國民必備的素養能力。本研究採用資料探勘技術，擷取國內新聞語料及教科書文本中有關氣候變遷主題之關鍵詞彙，並透過視覺化呈現目標詞彙網絡，探討學生對詞彙網絡的理解。研究探討新聞語料詞彙網絡之特性，包含：1.巨量實徵：詞彙網絡為客觀之表徵，透過視覺化圖像呈現一目了然；2.探索關鍵：讓學生理解詞彙之間的關聯，探索重要詞彙意涵；3.創新思考：透過詞彙連結學習，突破既有傳統脈絡限制，有利於創新思維；4.批判反思：促進探究與反思科學新聞資料的內涵。有關文本關鍵詞彙之探查，以往係透過專家採人工方式逐一檢視，再透過討論審議取得共識。本研究突破查找限制，以計算語言學技術擷取關鍵詞彙，並透過多重驗證，構建工具信效度，可以作為關鍵詞彙探勘開放性工具雛形。研究成果期能提供教學輔助指引，及教科書編寫設計參考。 |
| 摘要(外文) | In the ever-changing technology, information explosion era, digital literacy is essential for everyone to effectively capture knowledge. This study uses data mining to extract the keywords in climate change appeared in the domestic news corpus and textbooks; then, the network of target words is visualized in order to explore students' understanding of the vocabulary network. The characteristics of the network of target words are as follows: 1. huge amount and empirical, 2. with critical concepts, 3. with creative thinking, 4. with critical reflection. First, the network of target words objectively represents the huge empirical data and is visualized for students to read information effectively. Second, the network of target words with critical concepts is easy for students to understand the relationship among phrases and explore the meanings of keywords. Third, it is helpful for students' creativity; they learn new associations by breaking the traditional context constraints. Fourth, the process above helps students reflect the connotation of scientific news. In the past, experts scrutinized the keywords of the text and reached the consensus on the keywords through discussion. This study breaks the restrictions on exploring information to capture keywords with computational linguistics techniques. Then, it builds the validity and reliability of the measurement instrument through multiple verification and provides the prototype of the open tool for exploring keywords. We expect the results can provide instructional guidance and textbook design as references. |
| 關鍵字 | 關鍵詞彙關鍵詞自動擷取關聯性分析 |
| 關鍵字(外文) | keywordsautomatic keyword extractionassociation rules |