Educational data mining (EDM) is the process of analyzing data obtained from schools, students and administrators. The data that is analyzed is obtained from computer information systems, such as test scores and attendance records. Data mining looks for patterns and associations to draw conclusions about performance and behavior.
Modern educational environments rely on technology to streamline operations and keep track of important student data. Software applications are also used to administer student lesson plans, facilitate the learning process and administer exams. Communication between students, teachers and parents is also becoming largely dependent upon Internet and computer technology. Educational data mining seeks to combine all of this data to uncover new insights.
Schools use insights from data mining to develop new learning programs, improve performance and address potential issues. The technique can be used to determine what conditions help students learn better or perform better on exams. Employing educational data mining has become so popular that worldwide conferences are regularly held to teach educators about the techniques and discover new ways of incorporating it into schools.
Some of the topics explored during educational data mining conferences include how to effectively use data mining, how to mine different sources of data, improvement methods for educational software, and how to interpret data mining results to improve classroom instruction. Just as marketers use data mining to uncover associations between consumer purchasing habits and marketing activities, educational data mining seeks to discover unspoken patterns of behavior. For example, educators could use it to determine the effectiveness of experimental forms of learning and performance feedback for high school students, such as self-directed learning and assessments based upon subjective written reviews rather than a letter grade.
Data mining is a way to gain insight into the minds of students and administrators, which may be difficult to uncover with direct research methods. Some colleges and universities may analyze the results of graduating students' performance on national standardized tests to monitor the quality of its classroom instruction. High scores in certain subject areas over others may indicate a need to adjust the method in which that material is delivered. Other learning tools besides the traditional lecture may be tried as a result of data mining.
For example, if data mining uncovers that students retain more information over time as a result of working on projects rather than multiple choice tests, educators may start implementing more projects in all classes. Data mining can also isolate how certain groups of students learn. Results of student performance may reflect trends among age groups and gender.
