Analisa Neuromarketing Berbasis Sinyal EEG Menggunakan Video Reklame Produk

Mas, Arbintoro (2025) Analisa Neuromarketing Berbasis Sinyal EEG Menggunakan Video Reklame Produk. Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 6022211001-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 April 2027. Download (5MB) | Request a copy

Abstract

Memahami keterlibatan konsumen sangat penting untuk mengoptimalkan strategi pemasaran. Electroencephalography (EEG) telah menjadi alat penting dalam penelitian neuromarketing, memungkinkan analisis objektif terhadap respons neural terhadap berbagai stimulus. Penelitian ini bertujuan untuk menganalisis aktivitas neural pada individu yang menonton iklan video yang dikategorikan sebagai "menarik" atau "biasa saja," serta mengeksplorasi hubungan antara pola neural dengan keterlibatan konsumen melalui beban kognitif dan perhatial visual.
Sebanyak 30 peserta sehat berusia 17 hingga 40 tahun direkrut untuk menonton 20 iklan video dengan tingkat daya tarik yang bervariasi. Data EEG direkam menggunakan Ultracortex Mark IV yang dilengkapi dengan Cyton 8 headset pada frekuensi sampling 250 Hz. Protokol eksperimen meliputi sesi menonton iklan dan kuesioner pasca-sesi untuk mengevaluasi memori, perhatian, dan keterlibatan emosional. Pipeline preprocessing yang kuat diterapkan untuk memastikan kualitas data, termasuk penyaringan bandpass (1–49 Hz), penghapusan artefak menggunakan Artifact Subspace Reconstruction (ASR), serta Independent Component Analysis (ICA) untuk koreksi artefak lebih lanjut. Sinyal yang telah dibersihkan kemudian didekomposisi ke dalam lima subband utama yaitu Delta, Theta, Alpha, Beta, dan Gamma.
Ekstraksi fitur dilakukan dengan mengukur atribut temporal dan frekuensi, termasuk Power Spectral Density (PSD), Mean Absolute Value (MAV), Standar Deviasi, Peak-to-Peak, Hjorth Activity, dan Zero Crossing Rate. Analisis difokuskan pada kombinasi kanal-subband tertentu seperti O1-Gamma, O2-Gamma, F7-Beta, dan F8-Beta. Hasil penelitian menunjukkan bahwa aktivitas Gamma pada kanal O1 dan O2 secara signifikan lebih tinggi untuk iklan "menarik," yang mengindikasikan peningkatan perhatian visual. Sebaliknya, aktivitas Beta pada kanal F7 dan F8 lebih tinggi untuk iklan "biasa saja," mencerminkan peningkatan beban kognitif.
Penelitian ini memberikan wawasan berharga tentang dinamika neural yang mendasari respons konsumen terhadap iklan. Dengan mengidentifikasi pola EEG yang unik terkait dengan tingkat keterlibatan iklan, temuan ini menekankan potensi EEG dalam mengembangkan strategi neuromarketing. Studi ini juga menyoroti kegunaan analisis fitur yang terarah dalam memahami preferensi konsumen, membuka peluang untuk pengembangan kampanye pemasaran yang lebih efektif. Dengan kontribusi ini, penelitian ini menetapkan kerangka kerja yang kuat untuk penerapan EEG dalam analisis perilaku konsumen baik secara praktis maupun akademis.
=====================================================================================================================================
Understanding consumer engagement is critical to optimizing marketing strategies. Electroencephalography (EEG) has become an essential tool in neuromarketing research, allowing objective analysis of neural responses to various stimuli. This study aimed to analyze neural activity in individuals watching video ads categorized as “interesting” or “ordinary,” and explore the relationship between neural patterns and consumer engagement via cognitive load and visual attention.
Thirty healthy participants aged 17 to 40 years were recruited to watch 20 video ads of varying levels of appeal. EEG data were recorded using an Ultracortex Mark IV equipped with a Cyton 8 headset at a sampling rate of 250 Hz. The experimental protocol included an ad-watching session and a post-session questionnaire to evaluate memory, attention, and emotional engagement. A robust preprocessing pipeline was implemented to ensure data quality, including bandpass filtering (1–49 Hz), artifact removal using Artifact Subspace Reconstruction (ASR), and Independent Component Analysis (ICA) for further artifact correction. The cleaned signal was then decomposed into five major subbands, which are Delta, Theta, Alpha, Beta, and Gamma.
Feature extraction was performed by measuring temporal and frequency attributes, including Power Spectral Density (PSD), Mean Absolute Value (MAV), Standard Deviation, Peak-to-Peak, Hjorth Activity, and Zero Crossing Rate. The analysis focused on specific channel-subband combinations such as O1-Gamma, O2-Gamma, F7-Beta, and F8-Beta. The results showed that Gamma activity in O1 and O2 channels was significantly higher for “interesting” ads, indicating increased visual attention. In contrast, Beta activity in F7 and F8 channels was higher for “ordinary” ads, reflecting increased cognitive load.
This study provides valuable insights into the neural dynamics underlying consumer responses to advertising. By identifying unique EEG patterns associated with levels of ad engagement, these findings highlight the potential of EEG in developing neuromarketing strategies. This study also highlights the usefulness of targeted feature analysis in understanding consumer preferences, opening up opportunities for developing more effective marketing campaigns. With this contribution, this study establishes a robust framework for the application of EEG in consumer behavior analysis both practically and academically.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	beban kognitif, EEG, iklan video, neuromarketing, perhatian visual, cognitive load, EEG, neuromarketing, video advertisements, visual attention
Subjects:	T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5102.9 Signal processing.
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Electrical Engineering > 20101-(S2) Master Thesis
Depositing User:	Arbintoro Mas
Date Deposited:	26 Jan 2025 07:53
Last Modified:	26 Jan 2025 07:53
URI:	http://repository.its.ac.id/id/eprint/116898

Actions (login required)

View Item