The Analysis Of E-learning Using Decision Tree

With the application and popularization of E-learning, the influence of learners' nonintelligence factors and E-learning behaviors should gain more attention and an in-depth study, to provide rationale for the teachers to make instruction designs and schedules, and to provide support and reference for teachers who can improve instructions and adjust strategies. In order to analyze the effect of these factors on E-learning and to make good use of substantial existing data and system records, this study classifies these data and makes an in-depth study on these data with the method of decision tree. After making use of the method, this paper constructs a prediction model of E-learning effect, which is based on learners' "non-intelligence factors" and "E-learning behaviors". Moreover, a forecasting mechanism of the effect of E-learning, which provides great assistance and supports for the teachers to optimize their teaching methods, is achieved by further integrating of the two aspects.From this study, it is found that, in one way, the E-learning effect is affected by non-intellectual factors such as self-efficiency, learning interest, ambiguity tolerance and the condition of getting touch with internet. In another way, the E-learning effect is also directly influenced by the performances of accomplishing homework, participating discuss and looking up notice, etc. After combining the two patterns together, this study constructs a prediction mechanism, which contains the pre-test forecasting pattern (based on the nonintellectual factors) and the real time feedback pattern (based on the E-learning behaviors). This mechanism not only can assist the teachers to make predictions on the learning outcomes that the learner might get, but also can be associated with the teaching strategies in teaching practice, thus further optimization of the teaching effect is achieved.