EDM is the method of applying the concepts of data mining in the education domain. It focuses on applying the data mining algorithms on the humungous data, from multiple sources. In the wake of the emergence and growth of e-learning, and the concepts like surveys and peer assessments, a lot of data about a student gets generated and not everything is analyzed optimally.

The idea here is to use clustering and relationship mining, in the schools itself. This should preferably be applied in the middle school stage. This can help in gathering the data about the preference pattern observed in a student or say the kind of stuff he/she picks up for peer assessment. This can give a lot of perspective on the interest area of the student. The algorithms can be extrapolated, to be applied to student, grade-wide, in order to group them together based on their preferences. Generally, the counselor, allotted to a student, is on the basis of the serial order. Rather than doing this, if the counselor is allotted based on these preference clusters, it would become relatively easy for both to communicate. The counselor can channelize the session to suit the interest of only one kind of group assigned to him/her, rather than having students with multifarious interests and then juggle from one to another. Similarly, for a student, the interaction becomes far more valuable, if he gets to talk to like-minded people, who have similar career goals and the counselor who speaks the language that they can relate to.

The relationship mining, in this case, can help identify the issues faced by such students. And they are going to be more or less similar to the same cluster of students. Hence, the counselor can provide better counseling by even talking to a set of students, rather than one on one interaction. This can take a lot of load off the shoulders of the counselors and they can find more time to have engaging interactions with their students.