The Research For Network Study Evaluation On The Basis Of Fuzzy Theory And Analytic Hierarchy Process

As the fast development of computer network technology, teaching on the platform of the network has become an important method for effective education. The biggest difference between network teaching and traditional teaching is that the network teaching tends to make students independent and initiative in the process of learning. How to examine the results of learning on the teaching network, and how to help teachers choose their online teaching contents by obtaining effective mutual feedbacks from students have become important tasks of network teaching evaluation system. However, presently, most network teaching evaluation systems are restricted on online tests. They are little different than paper tests, and thus can not be used to evaluate students'learning progress effectively and efficiently. The quality-oriented education in our country has emphasized that not only the evaluation for the learning results should be valued, but also the evaluation for the learning process. Their positive effects on teaching and learning should be valued. In addition, evaluations should run through the whole process of learning, rather than only being done at the end. Our country has not published a standard for the evaluation of network teaching. Therefore, it is necessary to study the methods for network learning evaluation, especially the formative evaluation, which features in the process of network learning.This thesis will construct an innovative network learning evaluation system, which tends to solve the problem that present ineffective network teaching evaluation systems only use test scores for the learning process but do not pay attention to the evaluation . The thesis will start from analyzing the current situation, requirements, and characteristics of students in network learning. Then, it will create an evaluation system and improve the indexes. After that, a learning level evaluation model, which is suitable for the network learning evaluation system, as well as an evaluation system, which consists of quantization and non-quantization, will be proposed. We will use data mining technology to extract useful information from daily studying activities recorded on the web log, especially the key information that will affect students'study results, and use it as the source for learning process evaluation. Then, we will use evaluation information to build up a three-dimensional data cube, which will enable teachers to observe students'learning from every aspect. The system will first automatically collect the frequencies and the time that the students spend on learning, calculate the students'learning ability index, propose an evaluation approach that iscombined with fuzzy theory and analytic hierarchy process, and reduce human factor by making evaluation weights through analytic hierarchy process. Finally, we will apply fuzzy evaluation to the accumulated learning process evaluation information, and export the results in scores or fuzzy grades, which would help teachers to get overall, comprehensive and objective students'network learning evaluation.Being proved in practice, the method, which uses fuzzy theory and analytic hierarchy process to evaluate students'overall achievement, will provide objective and comprehensive information to help teachers to completely know how students learn online. Although, to evaluate how well the students'master of knowledge objectively and completely will increase examinations, the evaluation only needs to be graded fuzzily, such as excellent, good, and etc. Teachers can leave the calculations and statistical work to the computer, which is very easy to handle. This method will enhance the reliability of the data and the efficiency of network learning evaluation. The evaluation system which adopts the method proposed in the paper has obtained a good result in practical applications.