<![CDATA[Kidung Dewa Yadnya is a sacred Balinese vocal art that plays a vital role in Hindu religious ceremonies. However, its existence faces serious challenges in the modern era. Globalization and the dominance of popular culture among the younger generation tend to shift their attention away from local music and traditions due to a preference for modern genres better known through digital media. This puts traditional culture at risk of declining interest and sustainability without appropriate documentation or revitalization strategies. Therefore, this study aims to develop a classification system for Kidung Dewa Yadnya using digital signal processing and artificial intelligence approaches. The Mel-Frequency Cepstral Coefficients (MFCC) method is used to extract acoustic characteristics from sound signals because of its ability to represent human auditory perception. The MFCC process includes pre-emphasis, framing, windowing, fast fourier transform, mel filter bank, discrete cosine-transform, and cepstral lifting stages. The extracted feature vectors are used as input for the K-Nearest Neighbor (KNN) algorithm for classification. The research data were obtained from voice recordings of Kidung Dewa Yadnya singers in .wav format. The test results showed the highest accuracy of 81.25% at k = 1 and k = 2, with an average accuracy above 78%. This value indicates that the combination of MFCC and KNN methods is effective in recognizing the acoustic patterns of kidung. This research is expected to support the preservation of Balinese culture through the digitization of kidung and serve as a basis for the development of an artificial intelligence-based traditional music classification system.]]>
Copyrights © 2026
