PENGUKURAN BRAND STRENGTH LAPTOP AXIOO DAN ADVAN MENGGUNAKAN TEXT MINING DAN SEMANTIC RETRIEVAL
DOI:
https://doi.org/10.31539/0c8j0c78Abstract
Penelitian ini bertujuan menganalisis persepsi pengguna terhadap laptop Advan dan Axioo melalui komentar YouTube menggunakan text mining, analisis sentimen, dan semantic retrieval berbasis embedding. Data dikumpulkan dari 21 video dan menghasilkan 3.519 komentar, kemudian diperoleh 3.417 komentar unik setelah penghapusan duplikat. Pelabelan sentimen awal menggunakan TextBlob, sedangkan klasifikasi dilakukan dengan TF-IDF dan Multinomial Naive Bayes. Pembagian data dilakukan secara stratified 80:20 dan Random Oversampling hanya diterapkan pada data latih untuk mencegah data leakage. Hasil pengujian menghasilkan accuracy 0,730, weighted F1-score 0,782, dan macro F1-score 0,458. Distribusi label terdiri atas 250 positif, 3.050 netral, dan 117 negatif. Indeks brand strength berbasis sentimen sebesar 0,031429 untuk Advan dan 0,043860 untuk Axioo. Semantic retrieval digunakan untuk menemukan komentar yang relevan terhadap query.
Kata kunci: Brand strength, TF-IDF, Multinomial Naive Bayes, YouTube, semantic retrieval.
References
[1] O. Grljević dan Z. Bošnjak, “Sentiment analysis of customer data,” Strategic Management, vol. 23, no. 3, hlm. 38–49, 2018, doi: 10.5937/StraMan1803038G.
[2] L. Xinwei, Y. K. Tse, dan F. Fastoso, “Unleashing the power of social media data in business decision making: an exploratory study,” Enterprise Information Systems, vol. 18, no. 1, hlm. 2243603, Jan 2024, doi: 10.1080/17517575.2023.2243603.
[3] Institute of Economy and Informatics, Faculty of Organization and Management of Silesian University of Technology dan M. Wyskwarski, “SENTIMENT ANALYSIS OF COMMENTS POSTED ON YOUTUBE VIDEOS RELATED TO PHOTOVOLTAICS,” SPSUTOM, vol. 2023, no. 185, hlm. 617–637, 2023, doi: 10.29119/1641-3466.2023.185.36.
[4] S. Pichad, “Analysing Sentiments for YouTube Comments using Machine Learning,” IJRASET, vol. 11, no. 5, hlm. 1934–1938, Mei 2023, doi: 10.22214/ijraset.2023.51973.
[5] H. Jin, “From Information to Affirmation: An Investigation on the Echo Chamber Effect from YouTube Comments under Technology Product Reviews,” 2022, doi: 10.17605/OSF.IO/J2WQM.
[6] K. Victor Rajan dan F. Frejus, “Machine Learning Techniques for Opinion Mining from Social Media,” Computer, Mei 2022, doi: 10.54647/computer52271.
[7] Former Dean and Assistant Marketing Professor at Graduate School of Management at Sharif University of Technology, Iran, Dr. A. Miremadi, O. Ghanadiof, dan Graduate of Industrial Management, School of Technology, University of Central Missouri, USA, “The Ultimate Influences of Brand Equity Dimensions on Consumer Decision in Hi-Tech Market,” AJRSP, vol. 3, no. 27, hlm. 27–46, Jul 2021, doi: 10.52132/Ajrsp.e.2021.272.
[8] M. Ranjan, S. Tiwari, A. M. Sattar, dan N. S. Tatkar, “A New Approach for Carrying Out Sentiment Analysis of Social Media Comments Using Natural Language Processing,” dalam RAiSE-2023, MDPI, Jan 2024, hlm. 181. doi: 10.3390/engproc2023059181.
[9] R. Wati, S. Ernawati, dan H. Rachmi, “Pembobotan TF-IDF Menggunakan Naïve Bayes pada Sentimen Masyarakat Mengenai Isu Kenaikan BIPIH,” JAMIKA, vol. 13, no. 1, hlm. 84–93, Apr 2023, doi: 10.34010/jamika.v13i1.9424.
[10] N. Umar dan M. Adnan Nur, “Application of Naïve Bayes Algorithm Variations On Indonesian General Analysis Dataset for Sentiment Analysis,” J. RESTI (Rekayasa Sist. Teknol. Inf.), vol. 6, no. 4, hlm. 585–590, Agu 2022, doi: 10.29207/resti.v6i4.4179.
[11] A. S. Iedwan, N. Mauliza, Y. Pristyanto, A. D. Hartanto, dan A. N. Rohman, “Comparative Performance of SVM and Multinomial Naïve Bayes in Sentiment Analysis of the Film ‘Dirty Vote,’” sji, vol. 11, no. 3, hlm. 839–848, Nov 2024, doi: 10.15294/sji.v11i3.10290.
[12] P. Wibowo dan C. Fatichah, “An in-depth performance analysis of the oversampling techniques for high-class imbalanced dataset,” regist. j. ilm. teknol. sist. inf., vol. 7, no. 1, hlm. 63, Feb 2021, doi: 10.26594/register.v7i1.2206.
[13] S. F. N. Permatasari dan E. Ermatita, “A Two-Stage Hybrid Oversampling and Ensemble Learning Framework for Improved Type 2 Diabetes Mellitus Classification,” ijeeemi, vol. 8, no. 2, hlm. 202–217, Mei 2026, doi: 10.35882/ijeeemi.v8i2.308.
[14] D. Katta, “Multilingual Text Document Clustering and Classification,” IJRASET, vol. 13, no. 7, hlm. 1009–1014, Jul 2025, doi: 10.22214/ijraset.2025.73119.
[15] S. R. El-Beltagy dan M. A. Abdallah, “Exploring Retrieval Augmented Generation in Arabic,” 2024, arXiv. doi: 10.48550/ARXIV.2408.07425.
[16] Y. A. Singgalen, “Implementation of CRISP-DM for Social Network Analysis (SNA) of Tourism and Travel Vlog Content Reviews,” mib, vol. 8, no. 1, hlm. 572, Jan 2024, doi: 10.30865/mib.v8i1.7323.
[17] S. Magsarjav, M. Humphries, J. Tuke, dan L. Mitchell, “Best Preprocessing Techniques for Sentiment Analysis,” 2026, arXiv. doi: 10.48550/ARXIV.2606.24055.
[18] C. L. H. S. Nivak, “Sentiment Analysis of Ten Tourism Destinations in Indonesia,” J. Ind Tour Dev Std, vol. 9, no. 3, hlm. 162–168, Sep 2021, doi: 10.21776/ub.jitode.2021.009.03.03.
[19] N. Umar dan M. Adnan Nur, “Application of Naïve Bayes Algorithm Variations On Indonesian General Analysis Dataset for Sentiment Analysis,” J. RESTI (Rekayasa Sist. Teknol. Inf.), vol. 6, no. 4, hlm. 585–590, Agu 2022, doi: 10.29207/resti.v6i4.4179.
[20] S. Riyanto, I. S. Sitanggang, T. Djatna, dan T. D. Atikah, “Comparative Analysis using Various Performance Metrics in Imbalanced Data for Multi-class Text Classification,” IJACSA, vol. 14, no. 6, 2023, doi: 10.14569/IJACSA.2023.01406116.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Widi Anggraini, Lukman Sunardi, Cindi Wulandari, Budi Santoso
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
