Design And Implementation Of Multiple Teaching Evaluation System Based On Data Mining

As an important part of teaching work, teaching evaluations can not only test the effectiveness of teaching, but also promote the development of teaching. In recent years, because of multifaceted and multi-angle evaluation, diversified multi-level teaching evaluation can get the objective, comprehensive and accurate evaluation results, and effectively promote the progress of teaching, which has gradually become a hot research in the field of teaching evaluation.Because diversified teaching evaluation involves a lot of people and contents, its implementation is undoubtedly very complex. It would cost a lot of manpower and resources to implement evaluation through regular surveys, seminars, expert lectures and other forms, and effectiveness of evaluation would be hard to be guaranteed, at the same time, statistical analysis of evaluating the results will take a lot of time and evaluation results can not be timely feedback, impeding teaching improvement.With the continuous development of computer technology, especially the popularity of the Internet, online teaching evaluation is getting more and more popular. Online teaching evaluation can organize a large number of personnel to involve, getting the effective evaluation results which cannot be got by conventional methods, at the same time, it can analyze and process evaluation results using the powerful computing capability of the computer, timely feedback to the evaluation results to reviewers.However, because evaluation model is often fixed and single, the existing online teaching evaluation system can only support the summative assessment, but can not meet the diverse needs of the evaluation. Therefore, it is necessary to design and implement diverse online teaching evaluation system.In this paper, based on the theory of multiple teaching evaluation, diversified teaching evaluation model was established, and diversified teaching evaluation system was realized based on the model. The model was abstraction and generalization of diversified teaching evaluation, carrying out the relevant theories. The work is mainly as follows:(1) It researched and discussed the basic theory of data mining, analyzed background of data mining from two perspectives of business needs and technical, and especially clarified the basic concepts and characteristics of data mining. It analyzed several commonly available data mining methods, and gave the data mining architecture and process model. Finally, it pointed out the data mining application significance in university teaching, analyzed advantages and disadvantages of rough set theory, association rules and decision tree methods, and researched data mining of teaching evaluation, combining with tits characteristics and requirements.(2) It studied the theoretical basis of diversified teaching and evaluation, including basic concepts and classification. It analyzed the needs of diverse teaching evaluation, and pointed out its essence: diversity of teaching evaluation is a multi-level, multi-faceted, multi-angle evaluation, not only attaching importance to summative assessment, more emphasizing on the process of evaluation, stressing the evaluation feedback , incentives and guidance. Then, in the pluralistic basis of the theory of teaching evaluation, it constructed a variety teaching evaluation model, explicating evaluation relationship model, quantitative evaluation model, qualitative evaluation model, and process evaluation model in details. The model fully reflects the characteristics of the diverse teaching: diversification of the evaluation main body, the evaluation criteria, the assessment contents and the evaluation methods.(3) Because it is basic and necessary to establish a scientific and effective evaluation system in teaching evaluation, it is necessary to clarify the principles of the teaching evaluation index building before making the system overall design. In this thesis, the diversification of teaching evaluation system must follow the principles as follows: scientific principles, fairness principle, incentive principle, comprehensiveness principle, rationality principle and development principle.(4) Taking into account that the evaluation of teaching is not a single judgmental process but multi-dimensional, it analyzed to determine the main object of teaching evaluation and this paper choose the four rating main body: teachers, students, experts with peer authorities and teaching directors. Teacher self-evaluation is the understanding of the teachers themselves towards teaching quality. Students are the subject in teaching, so it is the direct approach to reflect the quality of teaching. Expert evaluation is the understanding of the experts towards teaching quality. The leadership's participating in the teaching evaluation can not only detect problems in teaching management timely, but also understand and master teachers'first-hand information directly, which is the important basis for teaching management decisions of the school.(5) It carried out the system overall design adopting C/S two-tier structure, gave the system's physical and logical functional structure, and designed teaching evaluation process based on data mining: identification of mining targets and goals, data acquisition, data conversion, data algorithm mining, regulations results analyzing after mining and application of knowledge. Finally, the system database was designed: clarifying database design process, constructing diversified teaching evaluation data model from three points of view of elective, evaluation and data processing, establishing the correspondent system back-end database tables according to system requirements analysis.(6) It realized the system, using Visual C # as front development tools and SQL Server 2000 as server-side database, and gave the specific running instance.(7) At last, system analysis and evaluation were carried out from the point of view of the application effects. After analysis, the application of the teaching evaluation system has the following advantages: good secrecy, convenient query, data backup and recovery, simple operation, convenient modification and high performance.