應用潛在語意分析增強學生的閱讀能力   全文下載

Using LSA-Based Tools to Enhance Students’ Chinese Reading Ability

本研究建立線上中文閱讀診斷與適性圖書推薦系統。系統內建204本電子書，每本書均有中文閱讀診斷測驗，共計5,456題。藉由診斷測驗評估學生的閱讀狀況，系統能依據診斷測驗結果，結合潛在語意分析，產生建議閱讀圖書清單供學生進行閱讀。本研究以準實驗研究法設計實驗流程，4年級兩個班為實驗組，兩個班為控制組，共有100位學生參與。實驗組閱讀線上中文閱讀診斷與適性圖書推薦系統產生的建議閱讀圖書清單，控制組依自己的喜好挑選圖書閱讀。經過一學期的實驗後，發現實驗組學生的中文閱讀理解測驗總分、語意能力、語法能力、比較分析，以及推論能力與控制組相比具有顯著性差異，顯示線上中文閱讀診斷與適性圖書推薦系統成效良好。

The features of tablet personal computer are high portability, support for multi-finger touch input, lower price, and multi communication interface. In this research, latent semantic analysis (LSA) and tablet personal computer (TPC) were applied to enhance students’ reading ability. Instructors applied LSA-based tools which expand students’ vocabulary learning material and generated recommendation story books. LSA-based Chinese vocabulary learning system and reading comprehension courses are applied with TPCs at school. After an online diagnostic test, the system generated expanded vocabularies and recommended story books according to students’ test result. Students learned the expanded vocabularies and read recommendation story books on their TPC. After 9 weeks of experiment, a Chinese reading comprehension test, a learning achievement test were conducted to measure the variation. Finally, 4 out of 7 abilities in the reading comprehension test were improved.