利用文章內容分析輔助學習歷程探勘

由於網路的蓬勃發展，線上教學已成為眾多輔助老師教學方式中，最主要的一種。線上學習有著不限時間、地點的特性，學生的線上學習行為如：上線次數、上線時間、閱讀教材的內容…等數據皆有學習歷程紀錄，因此有許多研究投入對學生線上活動及其學習成效之關係探討。然而，線上行為的紀錄中累計型的數據例如：上線時間、上線次數…等並無法表現出學生在各觀念或是主題的學習成果，僅能代表學生於此平台上活動的時間長短；但因學習態度及學生資質不同，時間長並不能代表學習成效佳，時間短也未必學習不足。因此本研究著重於學生在線上活動中所發表的文章內容及相關數據之分析，希望藉由分析結果來預測學生學習不足的觀念或主題。本研究的方式是將學生所發表的文章視為是對各觀念或主題的學習成果，將文章進行兩大部分的分類，1.針對文章的內容以文字探勘及統計等運算方式評比出客觀的數值以表現文章的好壞；2.對文章所代表的觀念或主題進行分類，以找出相同或類似內容的文章相互比較，並分析各文章發表人之間的學習成效，以找出學習不佳者。老師亦可利用本系統所提供之介面，根據不同需求設定參數，及時修正對學生學習的分析。或利用本系統資料庫中對各時間點的分析資料與現在狀態的比較找出不同年度學生學習的差異，輔以補救教學教材或是提供更多學生學習較弱的主題的內容來補強。本研究也利用現有的累積數據，對學生未來的學習成效進行預測，可提供老師可能學習失敗的學生名單，以便提早預防及提醒學生加強個人學習，使輔助學習的成效上更加完整。

關鍵字

輔助學習；主題分類；線上教學；文字探勘

並列摘要

Because of the flourishing development of the internet, online teaching has already become an important way of teacher's teaching. Online learning has not limited the time and place of the characteristics of students such as: login times, studying time, reading what kind of materials, such as data ... there are records of learning portfolios. There are lots of research has put into the online activities and explore the relationship between the effectiveness of learning. However, the records of online behavior data in a cumulative-type such as: login times, studying time …can only represent student's time accident of the activity on this platform; Because of the attitude of study and student qualification are different to each students, To stay for a long time can not represent the best learning effect. So our research focused on the activities of students in the online publication of the article content and related data analysis, and hope to predict with the analysis result that students study insufficient idea or theme. Our research is published by the students is considered an article on the concept or the theme of learning effect, and classification of two parts: 1. For the content of the articles in text mining and statistical algorithms, give each article a rating. 2. To represent the concept of the article or topic classification and find other articles which has similar topic or concept. The teachers can also utilize the interface that this system offered, establish the parameter according to different demands, revise the analysis of studying about students in time. Or using the database to analyzing or compare the difference of every years data. Our research can predict the learning effect by using the cumulative data that got in past years. In order to prevent and remained the students, we can make a student list that might be learn to fail in some concept. Teachers can use the list to give them more help.

並列關鍵字

Assisting studying ； Concept classification ； Text mining ； Online learning

參考文獻

鄭宇廷，應用學習成就及適性化教材之合作學習研究。中原大學資訊工

王嬈媚，網路學習歷程分析應用於提升學習動機之研究，2008

C.C. Chang, A study on the evaluation and effectiveness analysis of web-based learning portfolio (WBLP). British Journal of Educational Technology. 32, 4, 435-458. 2001.

C. Santamaría, J. Gonzalo and F. Verdejo, Automatic association of Web directories with word senses. Computat. Ling. 29, 3, pp.485–502. 2003.

C. Tang, S. Dwarkadas, and Z. Xu, On Scaling Latent Semantic Indexing for Large Peer-to-Peer Systems, Proceedings of the ACM SIGIR'04, 2004.

國際替代計量

利用文章內容分析輔助學習歷程探勘

全文下載

主題瀏覽