Fikri, Ahmad Aulia Zakiyal (2024) Pemodelan Analisis Sentimen Terhadap Review Pengguna Aplikasi Detikcom dengan Metode Naive Bayes Classifier. Other thesis, Institut Teknologi Sepuluh Nopember.
Text
5010201160-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2026. Download (4MB) | Request a copy |
Abstract
Detikcom merupakan salah satu pelaku utama dalam industri media digital di Indonesia dengan Tingkat akuisisi pasar sebanyak 65%. Terlepas dari berbagai prestasi yang diperoleh, Detikcom memiliki beberapa kekurangan yang didasarkan pada review pengguna di platform Google Play Store. Diketahui bahwa rating bintang 1 dari Detikcom (13%) merupakan salah satu yang tertinggi dibandingkan dengan kompetitornya. Hal ini secara tidak langsung menunjukkan tingkat user satisfaction yang rendah di sebagian pengguna sehingga akan berpengaruh pada jumlah page views dari aplikasi Detikcom. Berangkat dari permasalahan tersebut, penelitian ini berfokus untuk membangun model yang dapat mengidentifikasi sentimen pada pengguna Detikcom berdasarkan review pengguna pada platform Detikcom dengan metode Naïve Bayes Classifier serta mengidentifikasi faktor pendorong utama dari timbulnya setiap kategori sentimen dengan memanfaatkan Wordcloud yang telah diklasifikasikan berdasarkan masing-masing sentimen. Adapun tahap pemodelan yang akan dilakukan adalah identifikasi permasalahan, data scrapping, pelabelan awal data, data cleaning, persiapan model, pengoperasian model, pengukuran performa, klasterisasi pada setiap sentimen, penerbitan WordCloud, diskusi dan pembahasan, serta pengambilan kesimpulan. Berdasarkan hasil pada model yang dikembangkan, kecenderungan sentimen pengguna Detikcom cenderung positif (67,16%) dengan test accuracy sebesar 87,84%. Adapun keberadaan sentimen positif didasari oleh kata kunci akurat, terpercaya, dan up to date, sedangkan sentimen negatif didasari oleh kata kunci iklan. Peneliti juga merekomendasikan PT Trans Digital Media untuk mengimplementasikan hasil penelitian ini sebagai faktor strengths dan weaknesses dalam proses perencanaan pengembangan produk aplikasi Detikcom.
==============================================================================================================================
Detikcom is one of the most powerful digital media service provider in Indonesia with 65% of market acquisition percentage. Apart from the various achievements obtained, Detikcom has several shortcomings based on user reviews on the Google Play Store platform. It is known that Detikcom's 1-star rating (13%) is one of the highest compared to its competitors. Departing from this problem, the author took the initiative to build a model that can identify sentiment tendencies among Detikcom users based on user reviews on the Detikcom platform using the Naïve Bayes Classifier method and identify the main driving factors for the emergence of each sentiment category by utilizing Wordcloud which has been classified based on each sentiment. The research stages that will be carried out are problem identification, data scrapping, initial data labeling, data cleaning, model preparation, model operation, performance measurement, WordCloud publishing, discussions, and concluding. Through this research processes, the output to be achieved is a Naïve Bayes Classifier Sentiment Analysis Model with subjectively standardized accuracy (90%) and precision level (95%) and WordCloud algorithm according to the identified sentiment. According to researches and model developed by the researcher, user review sentiments are mostly positive (67,16% of testing data) with the test accuracy score of 87,84%. According to the result, the researcher also found that user reviews with positive sentiment are most likely contain keywords “accurate (akurat)”, “trusted (terpercaya)”, and “up to date”, while the negative reviews are most likely contain keyword “ads (iklan)”. The researcher also recommend PT Trans Digital Media to implement the result of this research by considering the output as strengths and weaknesses in Detikcom apps development planning process.
Item Type: | Thesis (Other) |
---|---|
Uncontrolled Keywords: | Data Scrapping, Detikcom, Naïve Bayes Classifier, Sentiment Analysis, WordCloud, Analisis Sentimen, Data Scrapping |
Subjects: | Q Science > Q Science (General) > Q325.5 Machine learning. Support vector machines. |
Divisions: | Faculty of Industrial Technology and Systems Engineering (INDSYS) > Industrial Engineering > 26201-(S1) Undergraduate Thesis |
Depositing User: | Fikri Ahmad Aulia Zakiyal |
Date Deposited: | 19 Aug 2024 05:34 |
Last Modified: | 19 Aug 2024 05:34 |
URI: | http://repository.its.ac.id/id/eprint/109944 |
Actions (login required)
View Item |