語言:

基於FML之特徵相似性評估於日語/臺語語言學習應用 = FML-based Feature Similarity Assessment for Japanese/Taiwanese Language Learning Application

吳冠毅國立臺南大學資訊工程學系碩士班民105 [2016]

可在府城總館 2樓參考書區 (DC CSIE 105010 )取得(請點選下列選項)

傳送到

題名:
基於FML之特徵相似性評估於日語/臺語語言學習應用 = FML-based Feature Similarity Assessment for Japanese/Taiwanese Language Learning Application
著者: 吳冠毅
國立臺南大學資訊工程學系碩士班
主題: 卷積運算基因學習模糊推論知識本體模糊標記語言; ConvolutionalOperation GeneticLearning FuzzyInference AssessmentOntology FeatureSimilarity FuzzyMarkupLanguage
描述: 隨著地球村的發展以及網路的發達，人們更容易取得全世界的資訊，同時利用線上自主學習平台進行自我學習，也是現代教育的趨勢，例如: 可汗學院即為現在全球知名的線上學習平台。本論文以博幼英文單字教材為實驗來源，藉由建置學習教材知識本體，期能協助使用者進行英文單字的學習。首先，領域專家使用模糊標記語言(Fuzzy Markup Language, FML)建置模糊推論機制所需之知識庫(Knowledge Base, KB)及規則庫(Rule Base, RB)。接著，基於臺語和日語單字詞組間之念法相似性、撰寫相似性和文化相似性，推論臺語和日語單字詞組之容易學習度，再以基因學習最佳化知識庫及規則庫。此外，本論文將學習教材知識本體經由所提方法轉換為卷積運算所需之矩陣資料，以觀察學習教材中每個單字分類的變化與單字間的分布狀態。未來，本論文將進一步回饋卷積運算結果至臺語和日語單字詞組容易學習度機制，期望有助於使用者之英文單字的學習。.
Today, people are much easier to get the information in the world owing to the development of the global village and the popularity of the Internet. Meanwhile, it is a trend to learn through the online self-learning platform for the modern education. For example, Khan Academy is a famous online learning platform in the world. This thesis uses Boyo English-word teaching materials as the experimental data and we hope to help users learn English words via the constructed learning-material ontology. First, the domain experts use fuzzy markup language (FML) to construct the necessary knowledge base (KB) and rule base (RB) of the fuzzy inference mechanism. Then, we infer the easy-to-learn degree of the word pair between Japanese and Taiwanese based on their feature similarity in pronunciation, writing, and culture. After that, a genetic learning is also adopted to optimize the KB and RB. Additionally, in order to observe the variation in the classification and the distribution of the words, this thesis also uses the proposed approach to transform the constructed learning-material ontology into a matrix to execute the convolutional operation. In the future, this thesis will further feedback the results of the convolutional operation to the proposed learning mechanism between Japanese and Taiwanese. It is hoped to help users to learn English words..
出版者: 碩士論文--國立臺南大學資訊工程學系碩士班, 105學年度
建立日期: 民105 [2016]
格式: vi,99葉 : 圖 ; 30公分..
語言: 中文
資源來源: NUTN ALEPH

連結

返回到檢索清單

前一個記錄檢索清單 4 下ㄧ個記錄下ㄧ頁

基於FML之特徵相似性評估於日語/臺語語言學習應用 = FML-based Feature Similarity Assessment for Japanese/Taiwanese Language Learning Application

吳冠毅 國立臺南大學 資訊工程學系碩士班 民105 [2016]

正在檢索遠程資料庫，請稍等

吳冠毅國立臺南大學資訊工程學系碩士班民105 [2016]