Prediction of librarian interest in library management in Pamekasan with comparison of SVM and KNN algorithms

This study was conducted to determine the prediction of librarian interest in joining a library organization. Using survey data and interviews with librarians that produced 130 test data then divided into two groups of data, namely "interested" and "not interested". Using the Support Vector Machine (SVM) and K-Nearest Neighbors (KNN) models as a comparison of the performance of the two algorithms in classifying librarian interests. The results of the test data were then evaluated using a confusion matrix to assess the accuracy, precision, and recall of each model. The results of the interest predictions tested showed that the use of the SVM model was more consistent in classifying librarian interests with high accuracy, although there were some errors in the "Not Interested" category. While the results of interest predictions using the KNN model tended to dominate the prediction of the "Interested" category, there were more errors in identifying the "Not Interested" category. Both models show their respective advantages and disadvantages in classifying librarian interest predictions. From the results of this study, it can be a picture and insight into the effectiveness of using the two models in classifying librarian interest predictions in joining a library organization and as a guide in choosing the right algorithm in similar research.