Application of association rule-based data mining algorithm in English teaching evaluation system

As a tool for science, technology, and communication, English is vital to the growth of colleges and universities. Improving the quality of English instruction has long been an important focus, where teachers and students are the primary participants. Teachers' attitudes, methods, and discipline significantly influence student achievement. This paper applies data mining techniques to university teaching assessment data to enhance evaluation objectivity. A method is proposed to minimise personal bias during data collection and to avoid subjective judgments. In data preprocessing, integration combines records from electronic and paper-based sources into a unified format, while discretisation converts continuous variables like cores, ages into categorical intervals, enabling rule-based algorithms such as Apriori. Using Apriori, associations among teaching indicators are mined, revealing strong correlations between single indicators and significant combined effects of multiple indicators. These results support suggestions for improving instruction based on identified relationships. By leveraging data mining, teacher evaluations become more personalised and objective, helping instructors refine teaching methods and better understand student needs. With the rapid growth of educational data, such as millions of records in institutional databases, this study demonstrates a practical application of association rule mining to extract valuable teaching insights and support effective management.