Melola Discovery

← Melola Discovery

Open access Jurnal SINTA 2

Analisis Sentimen Terhadap Kebijakan SubsidiPembelian Kendaraan Bertenaga Listrik Di Indonesia Menggunakan Pendekatan Inset Lexicon dan Metode Support Vector Machine

Jurnal Teknologi Informasi dan Ilmu Komputer (JTIIK) · Vol. 12 No. 6 (2025) · pp. 1303–1314

DOI
10.25126/jtiik.2025126
ISSN
2528-6579, 2355-7699
Published
2025-12-17
Publisher
Fakultas Ilmu Komputer, Universitas Brawijaya
Language
id
License
creativecommons.org/licenses/by-sa/4.0

Abstract

Bahasa Indonesia

Analisis sentimen berbasis leksikon merupakan metode yang umum digunakan untuk mengidentifikasi opini masyarakat terhadap isu-isu publik melalui media sosial. Penelitian ini membandingkan performa dua pendekatan leksikal, yaitu InSet lexicon dan Vader lexicon, dalam klasifikasi sentimen terhadap opini masyarakat mengenai kebijakan subsidi kendaraan listrik di Indonesia. Proses pelabelan sentimen dilakukan secara otomatis menggunakan masing-masing leksikon, kemudian diklasifikasikan menggunakan algoritma Support Vector Machine (SVM) dengan dua pendekatan representasi fitur, yaitu TF-IDF dan Word2Vec. Hasil klasifikasi menunjukkan bahwa InSet lexicon menghasilkan distribusi sentimen negatif yang dominan dengan akurasi klasifikasi sebesar 71%, sedangkan Vader lexicon lebih banyak mengidentifikasi sentimen positif dengan akurasi sebesar 64%. Evaluasi performa dilakukan menggunakan metrik akurasi, precision, recall, dan F1-score. Selain itu, visualisasi wordcloud digunakan untuk mengidentifikasi kata-kata kunci yang paling sering muncul dalam opini masyarakat, seperti “pemerintah”, “BBM”, dan “subsidi”, yang secara leksikal bersifat netral namun dapat membentuk arah sentimen tergantung konteks kalimat. Penelitian ini menunjukkan bahwa pemilihan leksikon dan representasi fitur berpengaruh signifikan terhadap hasil klasifikasi, serta menegaskan pentingnya validasi pelabelan dan pengembangan leksikon berbasis domain dalam analisis sentimen kebijakan publik. Abstract Lexicon-based sentiment analysis is a commonly used method to identify public opinion on policy issues through social media. This study compares the performance of two lexical approaches, namely InSet lexicon and Vader lexicon, in classifying sentiment toward public responses to Indonesia’s electric vehicle subsidy policy. Sentiment labeling was performed automatically using each lexicon, followed by classification using the Support Vector Machine (SVM) algorithm with two feature representation techniques: TF-IDF and Word2Vec. The results show that InSet lexicon yielded a dominant distribution of negative sentiment with a classification accuracy of 71%, while Vader lexicon identified more positive sentiments with an accuracy of 64%. Performance evaluation was carried out using accuracy, precision, recall, and F1-score metrics. In addition, wordcloud visualization was used to identify the most frequently appearing keywords in public opinion, such as “pemerintah” (government), “BBM” (fuel), and “subsidi” (subsidy), which are lexically neutral but may carry sentiment depending on contextual use. This study highlights the significant impact of lexicon choice and feature representation on classification performance and emphasizes the importance of label validation and domain-specific lexicon development in sentiment analysis for public policy evaluation.

English

Analisis sentimen berbasis leksikon merupakan metode yang umum digunakan untuk mengidentifikasi opini masyarakat terhadap isu-isu publik melalui media sosial. Penelitian ini membandingkan performa dua pendekatan leksikal, yaitu InSet lexicon dan Vader lexicon, dalam klasifikasi sentimen terhadap opini masyarakat mengenai kebijakan subsidi kendaraan listrik di Indonesia. Proses pelabelan sentimen dilakukan secara otomatis menggunakan masing-masing leksikon, kemudian diklasifikasikan menggunakan algoritma Support Vector Machine (SVM) dengan dua pendekatan representasi fitur, yaitu TF-IDF dan Word2Vec. Hasil klasifikasi menunjukkan bahwa InSet lexicon menghasilkan distribusi sentimen negatif yang dominan dengan akurasi klasifikasi sebesar 71%, sedangkan Vader lexicon lebih banyak mengidentifikasi sentimen positif dengan akurasi sebesar 64%. Evaluasi performa dilakukan menggunakan metrik akurasi, precision, recall, dan F1-score. Selain itu, visualisasi wordcloud digunakan untuk mengidentifikasi kata-kata kunci yang paling sering muncul dalam opini masyarakat, seperti “pemerintah”, “BBM”, dan “subsidi”, yang secara leksikal bersifat netral namun dapat membentuk arah sentimen tergantung konteks kalimat. Penelitian ini menunjukkan bahwa pemilihan leksikon dan representasi fitur berpengaruh signifikan terhadap hasil klasifikasi, serta menegaskan pentingnya validasi pelabelan dan pengembangan leksikon berbasis domain dalam analisis sentimen kebijakan publik. Abstract Lexicon-based sentiment analysis is a commonly used method to identify public opinion on policy issues through social media. This study compares the performance of two lexical approaches, namely InSet lexicon and Vader lexicon, in classifying sentiment toward public responses to Indonesia’s electric vehicle subsidy policy. Sentiment labeling was performed automatically using each lexicon, followed by classification using the Support Vector Machine (SVM) algorithm with two feature representation techniques: TF-IDF and Word2Vec. The results show that InSet lexicon yielded a dominant distribution of negative sentiment with a classification accuracy of 71%, while Vader lexicon identified more positive sentiments with an accuracy of 64%. Performance evaluation was carried out using accuracy, precision, recall, and F1-score metrics. In addition, wordcloud visualization was used to identify the most frequently appearing keywords in public opinion, such as “pemerintah” (government), “BBM” (fuel), and “subsidi” (subsidy), which are lexically neutral but may carry sentiment depending on contextual use. This study highlights the significant impact of lexicon choice and feature representation on classification performance and emphasizes the importance of label validation and domain-specific lexicon development in sentiment analysis for public policy evaluation.

Keywords

Collaboration network of the authors

12 authors · 22 co-authorship links · 4 across different organisations

Hover or tap a name to highlight its co-authors. Double-click to open the author's page.

  • Universitas Trisakti
  • Universitas Halu Oleo
  • Affiliation unknown
  • Different organisations
  • Same organisation
  • Affiliation unknown
  • Thicker line = more joint articles · bigger circle = more articles
Show as table
Co-authorOrganisationJoint articles
Gatot Budi SantosoUniversitas Trisakti2
Fiqih Hana SaputriUniversitas Trisakti1
Leni MuslimahUniversitas Trisakti1
Raden Davin RizkiUniversitas Trisakti1
Hariati LestariUniversitas Halu Oleo1
Renni MeliahsariUniversitas Halu Oleo1
Ramadhan TosepuUniversitas Halu Oleo1
Fithria FithriaUniversitas Halu Oleo1
Sartiah YusranAffiliation unknown1

Affiliations come from OpenAlex; journal OAI-PMH metadata carries none. Authors without an OpenAlex ID are matched by name, which can merge or split people.

Metadata source: OAI-PMH (metadata jurnal) https://jtiik.ub.ac.id/index.php/jtiik/en/oai, first harvested 27 Okt 2026, last seen 27 Okt 2026.
Original metadata (oai_dc)
<oai_dc:dc xmlns="http://www.openarchives.org/OAI/2.0/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:oai_dc="http://www.openarchives.org/OAI/2.0/oai_dc/" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.openarchives.org/OAI/2.0/oai_dc/  http://www.openarchives.org/OAI/2.0/oai_dc.xsd">
	<dc:title xml:lang="id">Analisis Sentimen Terhadap Kebijakan SubsidiPembelian Kendaraan Bertenaga Listrik Di Indonesia Menggunakan Pendekatan Inset Lexicon dan Metode Support Vector Machine</dc:title>
	<dc:title xml:lang="en">Analisis Sentimen Terhadap Kebijakan SubsidiPembelian Kendaraan Bertenaga Listrik Di Indonesia Menggunakan Pendekatan Inset Lexicon dan Metode Support Vector Machine</dc:title>
	<dc:creator xml:lang="en">Pratiwi, Dian</dc:creator>
	<dc:creator xml:lang="en">Khoerani, Nurizka</dc:creator>
	<dc:creator xml:lang="en">Sari, Syandra</dc:creator>
	<dc:creator xml:lang="id">Pratiwi, Dian</dc:creator>
	<dc:creator xml:lang="id">Khoerani, Nurizka</dc:creator>
	<dc:creator xml:lang="id">Sari, Syandra</dc:creator>
	<dc:subject xml:lang="id">Vader lexicon</dc:subject>
	<dc:subject xml:lang="id">InSet lexicon</dc:subject>
	<dc:subject xml:lang="id">subsidi</dc:subject>
	<dc:subject xml:lang="id">Kendaraan listrik</dc:subject>
	<dc:subject xml:lang="en">analisis sentimen</dc:subject>
	<dc:subject xml:lang="en">InSet</dc:subject>
	<dc:subject xml:lang="en">Vader</dc:subject>
	<dc:subject xml:lang="en">SVM</dc:subject>
	<dc:subject xml:lang="en">subsidi kendaraan listrik</dc:subject>
	<dc:description xml:lang="id">Analisis sentimen berbasis leksikon merupakan metode yang umum digunakan untuk mengidentifikasi opini masyarakat terhadap isu-isu publik melalui media sosial. Penelitian ini membandingkan performa dua pendekatan leksikal, yaitu InSet lexicon dan Vader lexicon, dalam klasifikasi sentimen terhadap opini masyarakat mengenai kebijakan subsidi kendaraan listrik di Indonesia. Proses pelabelan sentimen dilakukan secara otomatis menggunakan masing-masing leksikon, kemudian diklasifikasikan menggunakan algoritma Support Vector Machine (SVM) dengan dua pendekatan representasi fitur, yaitu TF-IDF dan Word2Vec. Hasil klasifikasi menunjukkan bahwa InSet lexicon menghasilkan distribusi sentimen negatif yang dominan dengan akurasi klasifikasi sebesar 71%, sedangkan Vader lexicon lebih banyak mengidentifikasi sentimen positif dengan akurasi sebesar 64%. Evaluasi performa dilakukan menggunakan metrik akurasi, precision, recall, dan F1-score. Selain itu, visualisasi wordcloud digunakan untuk mengidentifikasi kata-kata kunci yang paling sering muncul dalam opini masyarakat, seperti “pemerintah”, “BBM”, dan “subsidi”, yang secara leksikal bersifat netral namun dapat membentuk arah sentimen tergantung konteks kalimat. Penelitian ini menunjukkan bahwa pemilihan leksikon dan representasi fitur berpengaruh signifikan terhadap hasil klasifikasi, serta menegaskan pentingnya validasi pelabelan dan pengembangan leksikon berbasis domain dalam analisis sentimen kebijakan publik.
 
Abstract
 
Lexicon-based sentiment analysis is a commonly used method to identify public opinion on policy issues through social media. This study compares the performance of two lexical approaches, namely InSet lexicon and Vader lexicon, in classifying sentiment toward public responses to Indonesia’s electric vehicle subsidy policy. Sentiment labeling was performed automatically using each lexicon, followed by classification using the Support Vector Machine (SVM) algorithm with two feature representation techniques: TF-IDF and Word2Vec. The results show that InSet lexicon yielded a dominant distribution of negative sentiment with a classification accuracy of 71%, while Vader lexicon identified more positive sentiments with an accuracy of 64%. Performance evaluation was carried out using accuracy, precision, recall, and F1-score metrics. In addition, wordcloud visualization was used to identify the most frequently appearing keywords in public opinion, such as “pemerintah” (government), “BBM” (fuel), and “subsidi” (subsidy), which are lexically neutral but may carry sentiment depending on contextual use. This study highlights the significant impact of lexicon choice and feature representation on classification performance and emphasizes the importance of label validation and domain-specific lexicon development in sentiment analysis for public policy evaluation.</dc:description>
	<dc:description xml:lang="en">Analisis sentimen berbasis leksikon merupakan metode yang umum digunakan untuk mengidentifikasi opini masyarakat terhadap isu-isu publik melalui media sosial. Penelitian ini membandingkan performa dua pendekatan leksikal, yaitu InSet lexicon dan Vader lexicon, dalam klasifikasi sentimen terhadap opini masyarakat mengenai kebijakan subsidi kendaraan listrik di Indonesia. Proses pelabelan sentimen dilakukan secara otomatis menggunakan masing-masing leksikon, kemudian diklasifikasikan menggunakan algoritma Support Vector Machine (SVM) dengan dua pendekatan representasi fitur, yaitu TF-IDF dan Word2Vec. Hasil klasifikasi menunjukkan bahwa InSet lexicon menghasilkan distribusi sentimen negatif yang dominan dengan akurasi klasifikasi sebesar 71%, sedangkan Vader lexicon lebih banyak mengidentifikasi sentimen positif dengan akurasi sebesar 64%. Evaluasi performa dilakukan menggunakan metrik akurasi, precision, recall, dan F1-score. Selain itu, visualisasi wordcloud digunakan untuk mengidentifikasi kata-kata kunci yang paling sering muncul dalam opini masyarakat, seperti “pemerintah”, “BBM”, dan “subsidi”, yang secara leksikal bersifat netral namun dapat membentuk arah sentimen tergantung konteks kalimat. Penelitian ini menunjukkan bahwa pemilihan leksikon dan representasi fitur berpengaruh signifikan terhadap hasil klasifikasi, serta menegaskan pentingnya validasi pelabelan dan pengembangan leksikon berbasis domain dalam analisis sentimen kebijakan publik.
 
Abstract
 
Lexicon-based sentiment analysis is a commonly used method to identify public opinion on policy issues through social media. This study compares the performance of two lexical approaches, namely InSet lexicon and Vader lexicon, in classifying sentiment toward public responses to Indonesia’s electric vehicle subsidy policy. Sentiment labeling was performed automatically using each lexicon, followed by classification using the Support Vector Machine (SVM) algorithm with two feature representation techniques: TF-IDF and Word2Vec. The results show that InSet lexicon yielded a dominant distribution of negative sentiment with a classification accuracy of 71%, while Vader lexicon identified more positive sentiments with an accuracy of 64%. Performance evaluation was carried out using accuracy, precision, recall, and F1-score metrics. In addition, wordcloud visualization was used to identify the most frequently appearing keywords in public opinion, such as “pemerintah” (government), “BBM” (fuel), and “subsidi” (subsidy), which are lexically neutral but may carry sentiment depending on contextual use. This study highlights the significant impact of lexicon choice and feature representation on classification performance and emphasizes the importance of label validation and domain-specific lexicon development in sentiment analysis for public policy evaluation.</dc:description>
	<dc:publisher xml:lang="en">Fakultas Ilmu Komputer, Universitas Brawijaya</dc:publisher>
	<dc:date>2025-12-17</dc:date>
	<dc:type>info:eu-repo/semantics/article</dc:type>
	<dc:type>info:eu-repo/semantics/publishedVersion</dc:type>
	<dc:format>application/pdf</dc:format>
	<dc:identifier>https://jtiik.ub.ac.id/index.php/jtiik/article/view/9548</dc:identifier>
	<dc:identifier>10.25126/jtiik.2025126</dc:identifier>
	<dc:source xml:lang="id">Jurnal Teknologi Informasi dan Ilmu Komputer; Vol 12 No 6: Desember 2025; 1303-1314</dc:source>
	<dc:source xml:lang="en">Jurnal Teknologi Informasi dan Ilmu Komputer; Vol. 12 No. 6: Desember 2025; 1303-1314</dc:source>
	<dc:source>2528-6579</dc:source>
	<dc:source>2355-7699</dc:source>
	<dc:source>10.25126/jtiik.2025126</dc:source>
	<dc:language>id</dc:language>
	<dc:relation>https://jtiik.ub.ac.id/index.php/jtiik/article/view/9548/1514</dc:relation>
	<dc:rights xml:lang="en">Copyright (c) 2025 Dian Pratiwi, Nurizka Khoerani, Syandra Sari</dc:rights>
	<dc:rights xml:lang="en">https://creativecommons.org/licenses/by-sa/4.0</dc:rights>
</oai_dc:dc>