Analisis Sentimen Terhadap STMIK Widya Cipta Dharma Menggunakan Pendekatan Lexicon

Assyifa, Meireza Siti (2025) Analisis Sentimen Terhadap STMIK Widya Cipta Dharma Menggunakan Pendekatan Lexicon. S1 Sistem Informasi thesis, STMIK Widya Cipta Dharma.

[img] Text
2141902-S1-Sistem Informasi.pdf
Restricted to Repository staff only
Download (4MB) | Request a copy
[img] Text
2141902-S1-Jurnal.pdf
Download (453kB)

Abstract

Penelitian ini bertujuan untuk melakukan analisis sentimen terhadap STMIK Widya Cipta Dharma menggunakan pendekatan berbasis lexicon (lexicon-based approach). Analisis sentimen dilakukan untuk mengukur dan memahami opini publik terhadap STMIK Widya Cipta Dharma yang diharapkan dapat menjadi evaluasi instansi tersebut dan pendukung keputusan strategis yang lebih baik. Penelitian ini dilakukan di STMIK Widya Cipta Dharma Samarinda. Metode pengumpulan data yang digunakan yaitu dengan studi pustaka dengn mempelajari literature-literatur lain yang berkaitan dengan materi penelitian. Dengan studi lapangan yaitu pengumpulan data yang dilakukan melalui ulasan Google STMIK Widya Cipta Dharma dengan metode scrapping. Dalam penelitian ini metode pengembangan sistem yang digunakan dalam penelitian ini adalah CRISP-DM (Cross-Industry Standard Process for Data Mining). Data yang dikumpulkan melalui ulasan Google Stmik Widya Cipta Dharma, dan setelah diproses menghasilkan 216 data. Proses pelabelan data dilakukan menggunakan tiga teknik, yaitu kamus lexicon InSet , Labelling AI dan pelabelan manual. Adapun hasil akhir dari penelitian ini yakni berupa website yang menyajikan informasi mengenai sentimen terhadap STMIK Widya Cipta Dharma. Hasil labelling manual menghasilkan 171 sentimen positif, 16 sentimen negatif dan 29 sentimen netral atau 79,16% sentimen positif, 7.40% sentimen negatif dan 13.43% sentiment netral. Sedangkan hasil labelling menggunakan kamus lexicon Inset menghasilkan 99 sentimen positif, 95 sentimen negatif dan 22 sentimen netral atau 45.83% sentiment positif, 43.98% sentiment negatif dan 10.18% sentiment netral. Untuk hasil labelling menggunakan ChatGPT menghasilkan 133 sentimen positif, 7 sentimen negatif dan 76 sentimen netral. Atau 61.57% sentiment positif, 3.24% sentiment negatif dan 35.18 % sentiment netral. Hasil akurasi antara labelling manual dengan labelling lexicon InSet dapat diketahui akurasi sebesar 43.98% sedangkan hasil akurasi menggunakan labelling manual dan ChatGPT sebesar 66.20%. ============================================================ This study aims to conduct sentiment analysis of STMIK Widya Cipta Dharma using a lexicon-based approach. The sentiment analysis is performed to measure and understand public opinion towards STMIK Widya Cipta Dharma, which is expected to serve as an evaluation for the institution and support better strategic decision-making. The research was conducted at STMIK Widya Cipta Dharma, Samarinda. The data collection methods used include a literature review by studying relevant literature related to the research material, as well as field research, which involves gathering data from Google reviews of STMIK Widya Cipta Dharma using scraping techniques. In this study, the system development method used is CRISP-DM (Cross-Industry Standard Process for Data Mining). Data collected from Google reviews of STMIK Widya Cipta Dharma was processed, resulting in 216 data points. Data labeling was performed using three techniques: the InSet lexicon dictionary, AI labeling, and manual labeling. The final result of this research is a website that presents information on the sentiment towards STMIK Widya Cipta Dharma. The manual labeling results show 171 positive sentiments, 16 negative sentiments, and 29 neutral sentiments, or 79.16% positive sentiments, 7.40% negative sentiments, and 13.43% neutral sentiments. The labeling results using the InSet lexicon produced 99 positive sentiments, 95 negative sentiments, and 22 neutral sentiments, or 45.83% positive sentiments, 43.98% negative sentiments, and 10.18% neutral sentiments. The labeling results using ChatGPT produced 133 positive sentiments, 7 negative sentiments, and 76 neutral sentiments, or 61.57% positive sentiments, 3.24% negative sentiments, and 35.18% neutral sentiments. The accuracy between manual labeling and InSet lexicon labeling is 43.98%, while the accuracy between manual labelling and ChatGPT labeling is 66.20%.

Item Type: Thesis (S1 Sistem Informasi)
Additional Information: Pembimbing 1 : Dr. Heny Pratiwi, S.Kom., M.Pd., M.TI Pembimbing 2 : Wahyuni, S.Kom., M.Kom
Uncontrolled Keywords: Analisis Sentimen, CRISP-DM, Lexicon InSet, STMIK Widya Cipta Dharma
Subjects: Q Science > QA Mathematics > QA76 Computer software
Divisions: Sistem Informasi
Depositing User: Ms Meireza Siti Assyifa
Date Deposited: 11 Feb 2025 06:05
Last Modified: 11 Feb 2025 06:05
URI: http://repository.wicida.ac.id/id/eprint/6082

Actions (login required)

View Item View Item
wicida repository is powered by EPrints 3 which is developed by the School of Electronics and Computer Science at the University of Southampton. More information and software credits.