Implementation K-Nearest Neighbour for Student Expertise Recommendation System
Ichsan Taufik (a*), Yana Aditia Gerhana (a), Ade Irpan Ramdani (a), Mohamad Irfan (a)
a) Informatic Engineering, State Islamic University Sunan Gunung Djati Bandung, Jalan A.H. Nasution No. 105, 40614, Indonesia
*ichsan[at]uinsgd.ac.id
Abstract
The ability of students to determine their chosen field of expertise is still subjective, many students choose the field of expertise because their classmates choose the field of expertise not by considering their abilities and interests. This research uses the KNN classification method to determine areas of expertise that are in accordance with student expertise. The KNN method was chosen because it is a method that uses supervised algorithms where the results of new query instances are classified based on the majority of the categories in the KNN whose purpose is to classify test data based on training data. This system was tested using the confusion matrix method and the results were 98.30% of the total student data sample of 30 people.
Keywords: expertise, KNN, confusion matrix, recommendation
Topic: Computer Science