<![CDATA[The Balinese rindik gamelan is a traditional Balinese music that has a function as a balih-balihan art or as entertainment among the people during the royal era and has the same function until now. The simple rindik gamelan technique will produce a harmonious and melodious composition of tones, so that it acts as a stimulus carrier that can affect psychology and mood for the listener. This condition is caused by stress levels and energy levels. According to the mood theory by Thayer, music with positive stress levels and low energy refers to a pleasant (calm) mood, while positive stress levels and high energy levels refer to a happy (happy) mood. This study applies a branch of data mining, namely Music Information Retrieval (MIR) by classifying gamelan rindik labels based on the mood of the K-Nearest Neighbor strategy method. The algorithm is suitable to be implemented in nonlinear data sets such as music/song data, because the class decision line generated by K-NN is flexible. The data used are pieces of music files in mono .wav format with a duration of 20 seconds, which are then classified using the K-NN algorithm by the system. The resulting output is in the form of a mood label, namely satisfaction or joy. The results of the test using the confusion matrix obtained the percentage of classification accuracy 81%, recall 90.5%, precision 76%, and F-score 82.6% to find the average harmonic precision and recall. When compared with tests with accuracy, the same accuracy results are obtained, namely 81% or the best accuracy at a value of k = 7 and the fastest processing time is 0.0258 seconds.]]>
Copyrights © 2022
