Rule-Based Natural Language Understanding Based on Fuzzy Evaluation of Teaching Quality

This paper proposes a kind of method based on the rule natural language understanding, it can carry on scientific evaluation in understanding bases of the teachers´ teaching quality. First of all creates the domain library and evaluation library based on the quality of teaching evaluation system, establishes a rule for each word in domain and evaluation library. After identifies each domain and evaluation word, we call the corresponding rules to carry on fuzzy computing. Finally obtains the evaluation results. This method can use text description to evaluate teaching quality, has solved the defect only used the questionnaire in the process of evaluating teachers. We can carry on teaching evaluation task in any languages. Experiments proved that the method was feasible and practical.