Tinjauan Sentimen dan Topik Opini Publik Terhadap DPR RI Berdasarkan Komentar Twitter/X Menggunakan Valence Aware Dictionary and Sentiment Reasoner dan Latent Dirichlet Allocation

Aminatuzzuhriyah, Siti (2026) Tinjauan Sentimen dan Topik Opini Publik Terhadap DPR RI Berdasarkan Komentar Twitter/X Menggunakan Valence Aware Dictionary and Sentiment Reasoner dan Latent Dirichlet Allocation. Bachelor thesis, Institut Teknologi Kalimantan.

[img] Text
10221014_cover.pdf
Download (231kB)
[img] Text
10221014_statement_of_authenticity.pdf
Download (471kB)
[img] Text
10221014_publishing_agreement.pdf
Download (418kB)
[img] Text
10221014_approval_sheet.pdf
Download (469kB)
[img] Text
10221014_preface.pdf
Download (462kB)
[img] Text
10221014_abstract_id.pdf
Download (472kB)
[img] Text
10221014_abstract_en.pdf
Restricted to Repository staff only until 4 January 2028.
Download (428kB) | Request a copy
[img] Text
10221014_table_of_content.pdf
Restricted to Repository staff only until 4 January 2028.
Download (535kB) | Request a copy
[img] Text
10221014_illustrations.pdf
Restricted to Repository staff only until 4 January 2028.
Download (508kB) | Request a copy
[img] Text
10221014_tables.pdf
Restricted to Repository staff only until 4 January 2028.
Download (498kB) | Request a copy
[img] Text
10221014_notations.pdf
Restricted to Repository staff only until 4 January 2028.
Download (424kB) | Request a copy
[img] Text
10221014_chapter_1.pdf
Restricted to Repository staff only until 4 January 2028.
Download (792kB) | Request a copy
[img] Text
10221014_chapter_2.pdf
Restricted to Repository staff only until 4 January 2028.
Download (994kB) | Request a copy
[img] Text
10221014_chapter_3.pdf
Restricted to Repository staff only until 4 January 2028.
Download (1MB) | Request a copy
[img] Text
10221014_chapter_4.pdf
Restricted to Repository staff only until 4 January 2028.
Download (4MB) | Request a copy
[img] Text
10221014_conclusions.pdf
Restricted to Repository staff only until 4 January 2028.
Download (489kB) | Request a copy
[img] Text
10221014_bibliography.pdf
Download (451kB)
[img] Text
10221014_enclosure.pdf
Restricted to Repository staff only until 4 January 2028.
Download (867kB) | Request a copy
[img] Text
10221014_paper.pdf
Restricted to Repository staff only until 4 January 2028.
Download (746kB) | Request a copy
[img] Text
10221014_presentation.pdf
Restricted to Repository staff only until 4 January 2028.
Download (2MB) | Request a copy

Abstract

Perkembangan media sosial menjadikan Twitter/X sebagai salah satu indikator utama opini publik terhadap kinerja institusi negara, termasuk DPR RI. Namun, opini yang tersedia berjumlah besar dan menggunakan bahasa informal sehingga sulit dianalisis secara langsung. Penelitian ini bertujuan menganalisis sentimen dan memodelkan topik untuk memetakan aspirasi masyarakat terhadap DPR RI secara komputasional. Data dikumpulkan melalui web scraping pada periode 2016-2025 dan menghasilkan 16.379 data awal. Tahap pra-pemrosesan meliputi seleksi data, pembersihan noise, normalisasi peregangan kata dan istilah tidak baku, tokenisasi, serta stemming, sehingga diperoleh 13.192 data. Analisis sentimen dilakukan menggunakan tiga pendekatan berbasis VADER, yaitu ISV-RSN (InSet Vader with Runtime Score Normalization), ISV-SLA (InSet Vader with Structural Lexicon Adaptation), dan VTB (Vader Translation Based). Hasil evaluasi terhadap ground truth menunjukkan bahwa ISV-SLA memiliki performa terbaik dalam mengklasifikasikan sentimen teks informal berbahasa Indonesia. Data bersentimen negatif selanjutnya dianalisis menggunakan algoritma Latent Dirichlet Allocation (LDA) untuk mengidentifikasi isu utama. Hasil optimasi menunjukkan bahwa konfigurasi enam topik menghasilkan nilai coherence score tertinggi. Keenam topik tersebut mencerminkan fokus ketidakpuasan publik, meliputi kebijakan dan sidang komisi, representasi wakil rakyat, dinamika politik dan legislasi, aktivitas reses dan kelembagaan, fungsi aspirasi daerah, serta isu korupsi dan elite politik. Penelitian ini menyimpulkan bahwa integrasi leksikon teradaptasi dan pemodelan topik probabilistik efektif dalam merepresentasikan pola opini digital serta isu strategis terkait kinerja legislatif.

Item Type: Thesis (Bachelor)
Subjects: H Social Sciences > H Social Sciences (General)
H Social Sciences > HA Statistics
J Political Science > JA Political science (General)
Q Science > Q Science (General)
Divisions: Jurusan Matematika dan Teknologi Informasi > Sistem Informasi
Depositing User: Siti Aminatuzzuhriyah
Date Deposited: 10 Jul 2026 01:38
Last Modified: 10 Jul 2026 01:38
URI: http://repository.itk.ac.id/id/eprint/26279

Actions (login required)

View Item View Item