ANALISIS BIG DATA SINYAL FACIAL SURFACE ELECTROMYOGRAPHY MENGGUNAKAN APACHE SPARK PADA KOMPUTER TERDISTRIBUSI UNTUK PENINGKATAN AKURASI IDENTIFIKASI POLA EMOSI

    Auziah Mumtaz, - and Liptia Venica, - and Dewi Indriati Hadi Putri, - (2025) ANALISIS BIG DATA SINYAL FACIAL SURFACE ELECTROMYOGRAPHY MENGGUNAKAN APACHE SPARK PADA KOMPUTER TERDISTRIBUSI UNTUK PENINGKATAN AKURASI IDENTIFIKASI POLA EMOSI. S1 thesis, Universitas Pendidikan Indonesia.

    Abstract

    Identifikasi emosi berbasis sinyal sEMG wajah menghadapi tantangan akibat volume data yang besar dan kompleksitas sinyal fisiologis. Penelitian ini mengembangkan pendekatan analisis big data menggunakan kerangka kerja Apache Spark dalam lingkungan komputasi terdistribusi untuk mengolah keseluruhan dataset EmgDataVR (±30,4 juta baris). Jenis penelitian ini adalah eksperimental dengan pendekatan mixed-method dan model iterative-incremental development (IID). Metode mencakup preprocessing terdistribusi (feature selection, row filtering, dan deteksi outlier), serta penerapan algoritma clustering K-Means dan Bisecting K-Means. Evaluasi dilakukan dengan metrik silhouette coefficient untuk menilai kualitas pemisahan cluster. Konfigurasi terbaik diperoleh dari Bisecting K-Means dengan dua cluster (silhouette = 0,8639). Pola hasil clustering mengindikasikan dua kondisi emosi berdasarkan tingkat arousal: low arousal (aktivitas otot rendah) dan high arousal (aktivitas signifikan pada Zygomaticus dan Orbicularis, sesuai AU12 dan AU6). Temuan ini membuktikan bahwa pendekatan komputasi terdistribusi dapat meningkatkan efisiensi pemrosesan, akurasi hasil, dan skalabilitas sistem identifikasi emosi berbasis sinyal sEMG wajah. ----- Facial emotion recognition based on surface electromyography (sEMG) signals faces significant challenges due to the large volume of data and the complexity of physiological signals. This study aims to develop a big data analysis approach using the Apache Spark framework in a distributed computing environment to process the entire EmgDataVR dataset (±30,4 million rows). This is an experimental research employing a mixed-method approach, developed under the iterative-incremental development (IID) model. The proposed method includes a distributed preprocessing pipeline (feature selection, null filtering, and outlier detection), followed by the application of two clustering algorithms: K-Means and Bisecting K-Means. The clustering performance is evaluated using the silhouette coefficient metric to measure the quality of cluster separation. The Bisecting K-Means configuration with two clusters yields the most optimal result with a silhouette score of 0,8639. The resulting cluster patterns indicate two primary emotional states based on arousal levels: low arousal (low muscle activation) and high arousal (strong activation of the Zygomaticus and Orbicularis muscles, corresponding to AU12 and AU6). These findings demonstrate that a distributed computing approach can significantly improve processing efficiency, result accuracy, and the scalability of emotion identification systems based on sEMG signals.

    Baca Full Text klik disini
    [thumbnail of S_MKB_2103554_Title.pdf] Text
    S_MKB_2103554_Title.pdf
    Download (667kB)
    [thumbnail of S_MKB_2103554_Chapter1.pdf] Text
    S_MKB_2103554_Chapter1.pdf
    Download (307kB)
    [thumbnail of S_MKB_2103554_Chapter2.pdf] Text
    S_MKB_2103554_Chapter2.pdf
    Restricted to Staf Perpustakaan
    Download (1MB)
    [thumbnail of S_MKB_2103554_Chapter3.pdf] Text
    S_MKB_2103554_Chapter3.pdf
    Download (693kB)
    [thumbnail of S_MKB_2103554_Chapter4.pdf] Text
    S_MKB_2103554_Chapter4.pdf
    Restricted to Staf Perpustakaan
    Download (2MB)
    [thumbnail of S_MKB_2103554_Chapter5.pdf] Text
    S_MKB_2103554_Chapter5.pdf
    Download (341kB)
    [thumbnail of S_MKB_2103554_Appendix.pdf] Text
    S_MKB_2103554_Appendix.pdf
    Restricted to Staf Perpustakaan
    Download (255kB)
    Official URL: https://repository.upi.edu/
    Item Type: Thesis (S1)
    Additional Information: https://scholar.google.com/citations?user=KONB4JMAAAAJ&hl=en&authuser=2&oi=ao ID SINTA Dosen Pembimbing: Liptia Venica: 6779029 Dewi Indriati Hadi Putri: 6720737
    Uncontrolled Keywords: sEMG, identifikasi emosi, big data, Apache Spark, clustering, K-Means, Bisecting K-Means sEMG, emotion recognition, big data, Apache Spark, distributed computing, clustering, K-Means, Bisecting K-Means
    Subjects: Q Science > Q Science (General)
    Q Science > QA Mathematics
    Q Science > QA Mathematics > QA75 Electronic computers. Computer science
    Q Science > QP Physiology
    T Technology > T Technology (General)
    T Technology > TK Electrical engineering. Electronics Nuclear engineering
    Divisions: UPI Kampus Purwakarta > S1 Mekatronika dan Kecerdasan Buatan
    Depositing User: Auziah Mumtaz
    Date Deposited: 15 Aug 2025 07:48
    Last Modified: 15 Aug 2025 07:48
    URI: http://repository.upi.edu/id/eprint/135571

    Actions (login required)

    View Item View Item