PREDIKSI KEPUASAN PELAYANAN PERPUSTAKAAN MENGGUNAKAN ALGORITMA DECISION TREE (C4.5)

Abstract

Prediction of library service satisfaction is important in determining library growth and progress. To determine which attributes are prioritized by library managers is very useful for improving library services going forward. The results of the analysis using the Confusion Matrix can be concluded that the accuracy value of the determinant attribute on Y '(2 types of assessment) is better than Y (5 type of assessment) with an accuracy value of 75.50%. Based on the analysis of the two decision trees on Y and Y ', the attributes that equally affect the satisfaction of library services are Environmental Comfort (X13), Officer Clarity (X3), Collection Fulfillment (X2) and Courtesy (X9). Based on the combination of the decision tree image of Y 'with the weighting value table of each attribute, it can be concluded that the Collection Fulfillment Attribute (X2) is the most influential attribute on the library service satisfaction decision.Keywords: Prediction, Library, Accuracy, Algorithm, Decision Tree (C4.5)