Audio Music Language Classification and Analysis Using TOPSIS

Abstract

Indian Language classification in songs using Topsis comes under the field of Music Information Retrieval (MIR). Recently it has become a sought-after field among researchers for its future potential in organizing music corpus on video streaming and music streaming platforms, and also other broader studies in the fields like behavioural science of how a community can be traced through its music tastes and preferences. Music is characterize d by various features related to instrumentation, repetitive rhythmic structures, and gradual changes in harmony. Identifying a language from songs is considered a challenging task. We inquire into various machine learning algorithms to complete out the task of identifying audio language from music audio clips which include s classifiers such as SVM, KNN, Random Forest , LDA, Linear Regression, Naive Bayes and more obtaining the best accuracy with 70%.