在一般的教學環境下，學生與老師之間的互動主要是依賴上課時的教課與問問題方式來達成；隨著網際網路的發達，以學習網站作為教學上的輔助，雖然讓學生可以不限時間與地點，只要透過網路，即可達成互動的目的，然而，老師與學生之間的互動，仍受限於學生主動地去了解、探索。
藉由手機簡訊的功能，老師可以不用等待學生主動來詢問問題才能與學生溝通，透過我們行動輔助學習系統的功能，老師可以主動push學習訊息，給予需要的學生，讓學生在被動的狀態下，依然可以得到學習訊息。
在使用了手機簡訊功能後，我們面臨到的問題是，何時傳送最為恰當，以及哪些內容該何時傳送才不會被排斥；我們的系統在偵測學生傳送時間與內容的喜好程度，並且使用association rule的方式，取得預測的規則。在測試結果中，除了因學生突發的事件，使得我們較無法掌握學生回應訊息距傳送時的間隔外，對於何時傳送、以及傳送內容是否喜歡以及學生個人的緊張狀態，約有70%的預測結果。
根據勒溫的心理緊張系統，維持人的緊張狀態，對於回憶、記憶以及趨使往目標，具有正向的力量。因此，我們探討了學生的緊張狀態與學習上的各項因素，了解緊張狀態的提昇，對於學習成效上的幫助。藉由情緒偵測模組，取得學生的緊張程度，我們可以在學生緊張程度低於設定值時，將訊息加入具有威脅或壓力性質高的特性。
在BBN分析上，雖然沒有找到影響緊張程度變化的因子，但藉由資料的觀察，我們發現成績不及格者，在面對計概作業、考試前，其緊張程度有昇高的狀況；而透過問卷，有67%的學生認為訊息對他們的緊張狀態具有影響，有74%的人若因訊息使其緊張程度提昇，會去處理訊息相關的事項，而且有76%的人願意在3小時以內去處理。

Under the general teaching environment, to achieve the interaction between students and teachers relies on materials-teach and question –asking in class. As the Internet grows up, students don’t have to consider when or where they communicate with others, just by the learning web site. But the interaction between students and teachers is limited to active exploring by students.
With the help of short message, teachers don’t need to communication with students just by waiting them asking questions actively. Teachers could actively push learning-messages to those students who need by our mobile learning aided system. Students could receive learning-messages even in passive status.
After we get the function of short message, we have some problems about when to send and what to send is suitable. Our system gets the sending rules from the association rules trained by detecting what content they like and when to send they prefer. Besides some unexpected situation that the interval from the message we sent to the acknowledge students response we couldn’t accurately predict, we have about 70 percent accurate rate on predicting when to send, what content they like, and the level of nervous before/after they receive messages.
According to Lewin’s nervous system, it’s positive for someone on recall, memorization, or goal-approach by keeping one’s nervous status. Thus, we want to realize the relation between nervous status and other learning related attributes to improve learning performance. By getting students’ nervous level from affective detection module, we could change the message to become threatened or pressed while the students’ nervous level is below the threshold.
Though we didn’t find the factor affecting the changes on nervous, we found that the lowest-degree students’ nervous level grows up before the course test or the deadline of the homework according to observing the data. And we found 67 percent of the students who think that the message would make an effect on their nervous level, 74 percent of the students who would deal with the message if it levels up their nervous status, and 76 percent of the students who would deal with the message in 3 hours.

[1]申荷永, 充滿張力的生活空間, 貓頭鷹出版社
[2]李?吟,張德榮,洪寶蓮. 大學生學習與讀書策略量表, 中國行為科學社
[3]詹志禹, 賴世培. 應用統計. 國立空中大學.
[4]Bing Liu, Wynne Hsu, Yiming Ma. Integrating Classification and Association Rule Mining, KDD ‘98
[5]BKD (Bayesian Knowledge Discoverer) developed by Knowledge Media Institute of The Open University at http://kmi.open.ac.uk/projects/bkd.
[6]Deutsch. Field theory in social psychology. In G. Lindzey & Aronson, The Handbook of Social Psychology. New York. Vol 1. 1968
[7]Ekman, Paul. Are there basic emtions?, Psychological Review, 99(3):550-553. 1992.
[8]Elliot Soloway. Center for Highly Interactive Computing in Education at the University of Michigan at http://www.handheld.hice-dev.org/index.htm
[9]Elliot Soloway, Cathleen Norris, Phyllis Blumenfeld, Barry Fishman, Joseph Krajcik, and Ron Marx. Handheld Devices are Ready-at-Hand. Communications of the ACM. Vol. 44, No. 6. June 2001.
[10]Elliot Soloway, Wayne Grant, Robert Tinker, Jeremy Roschelle, Mike Mills, Mitchell Resnick, Robbie Berg, and Michael Eisenberg. Science in the Palms of Their Hands. Communications of the ACM. Vol. 42, No. 8. Augest 1999.
[11]Fayyad, U.M., Gregory, P.S., Smyth, P. and Uthurusamy, R. A Statistical Perspective on Knowledge Discovery in Database. Advances in Knowledge Discovery and Data Mining, 1996.
[12]Jeremy Roschelle and Roy Pea. A walk on the WILD side: How wireless handhelds may change CSCL, CSCL 2002
[13]Jiawei Han, Micheline Kamber. Data Mining: Concepts and Techniques. Morgan Kaufmann Publishers. USA. 2000.
[14]Lewin K. Field Theory in Social Science. (Ed. By D. Cartwright), New York: Harper & Brothor Publishers. 1951
[15]Lewin K. The Principles of Topological Psychology. New York: McGraw-Hill Book Company, Inc. 1936
[16]Mealey, D.L. Test Review Learning and study strategiees Inventory. Journal of Reading V31 N4 P382-85.
[17]Leeper, R. Lewin’s Topological and Vector Psychology. New York: University of Oregon. 1943
[18]Leidelmeijer, K., Emotions: An Experimental Approach. Tilburg University Press, 1991.
[19]Norman W. Paton and Oscar Diaz. Active Database, ACM computing Surveys, Vol. 31, No. 1, March 1999.
[20]Picard, R. W., Affective Computing. Cambridge, Mass.: MIT Press 1977.
[21]Plutchik, R. ‘A general psychoevolutionary theory of emotion,’ in Emotion Theory, Research, and Experience(R. Plutchik and H. Kellerman, eds.), vol. 1, Theories of Emotion, Academic Press, 1980.
[22]Robert J.Sternberg. Thinking Styles, Cambridge University Press. 1997
[23]Rosalind Picard, Barry Kort, and Rob Reilly. Project Summary: Exploring the role of Emotion in Propelling the SMET Learning Process. 2002.
[24]Tom M. Mitchell, Machine Learing, McGRAW-Hill International Editions, 1997
[25]Weinsten, C.E. Learning and Study strategies Inventory (LASSI), User’s Manual, 1987.