Analisis Sentimen Opini Publik Menggunakan Metode Naive Bayes Classifier(NBC)(Studi Kasus Review Smartphone Merk "A" Dan "B")

Syah, Mochammad Rozy Andrean (2021) Analisis Sentimen Opini Publik Menggunakan Metode Naive Bayes Classifier(NBC)(Studi Kasus Review Smartphone Merk "A" Dan "B"). Undergraduate thesis, Politeknik Negeri Jember.

	Text (Abstract) 1. Moch. Rozy Andrean S. - Abstract.pdf - Submitted Version Available under License Creative Commons Attribution Share Alike. Download (194kB)
	Text (Bab 1 Pendahuluan) 2. Moch. Rozy Andrean S. - Bab 1.pdf - Submitted Version Available under License Creative Commons Attribution Share Alike. Download (134kB)
	Text (Full Teks) 3. Moch. Rozy Andrean S. - Full Text.pdf - Submitted Version Restricted to Registered users only Download (3MB) | Request a copy
	Text (Daftar Pustaka) 4. Moch. Rozy Andrean S. - Daftar Pustaka.pdf - Submitted Version Available under License Creative Commons Attribution Share Alike. Download (231kB)

Abstract

Along with the development of time and technology, mobile phones that were originally only for communicating and sending short messages have developed into smart phones or smartphones. In Indonesia, there are many smartphone brands circulating, smartphone brands "A" and "B" smartphone development companies are competing to dominate the market share in Indonesia. To increase the market share of smartphone companies with the "A" and "B" brands, it is necessary to research public opinion on these companies. The research conducted is an analysis of public opinion sentiment on the smartphone brands "A" and "B" obtained from social media Twitter. One way to get data on Twitter is by using Text Mining using the Twitter API for data retrieval. The data obtained is processed by Text Preprocessing and the data is classified using the Naïve Bayes Classifier method. The validation process for measuring system performance using Confusion Matrix get resulted in an accuracy value of 74.6%, precision 82.2%, recall 74%, and f-measure 77.8% for brand "A" and an accuracy value of 63.4%, precision 60.6%, recall 86%, and f-measure 77.1% for brand "B".

Actions (login required)

View Item