Sentiment Analysis of Public Figure News using Sentiment Lexicon and Machine Learning

Tsabit, Fitriana Zahirah (2023) Sentiment Analysis of Public Figure News using Sentiment Lexicon and Machine Learning. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 05111942000011-Non_Degree.pdf - Accepted Version Restricted to Repository staff only until 1 October 2025. Download (2MB) | Request a copy

Abstract

Dalam dunia politik maupun dunia bisnis penilaian yang dibentuk media dapat menjadi sebuah citra dari seseorang. Citra seseorang merupakan komponen penting bagi seorang publik figur untuk mendapatkan popularitas dan keuntungan secara finansial. Saat ini dunia informasi berkembang dengan cepat sehingga lebih mudah mendapatkan informasi melalui situs berita online. Di Indonesia salah satu situs beritaonline terbesar yaitu Detik.com. Situs berita online menyediakan informasi dalam berbagai bentuk (teks, video atau gambar) dengan membaginya sesuai dengan topik-topik tertentu. Sehingga, dapat dilakukan sentimen analisis untuk mengetahui citra publik figur dari situs berita online melalui penerapan metode machine learning yaitu klasifikasi. Data yang digunakan berasal dari situs berita online yang diambil dengan cara scraping. Tahapan yang dilakukan preprocessing, gabungan score feature extraction menggunakan kamus lexicon, InSet, SentIl, Emolex dan juga terhadap gabungan kamus EmoTil (Emolex dan SentIl), EmoSet (Emotil dan InSet) dan SenSet (InSet dan SentIl) dengan TFIDF. Hasil dari penggabungan ini digunakan untuk klasifikasi machine learning dengan algoritma SVM dan Logistic Regression. Model yang telah dibangun akan dievaluasi dengan membandingkan nilai akurasi dari cross validation. Hasil akurasi terbesar dan terkecil akan dilakukan penembahan parameter untuk meningkatkan hasil dari rata-rata cross validation. Hasil dari akurasi terbesar adalah SVM dengan selisih 1.2% dengan Logistic Regression
=================================================================================================================================
In the world of politics and business, the judgment formed by the media can become an image of a person. A person's image is essential for a public figure to gain popularity and financial benefits. The world of information is developing rapidly, making it easier to get information through online news sites. In Indonesia, one of the largest online news sites is Detik.com. Online news sites provide information in various forms (text, video, or images) by dividing it according to specific topics. Thus, an analysis of sentiment can be carried out to determine the image of public figures from online news sites through machine learning methods, namely classification. The data used comes from online news sites taken by scraping. The stages carried out are preprocessing, combined score feature extraction using lexicon dictionaries, InSet, SentIl, and Emolex and also against the combined dictionaries EmoTil (Emolex and SentIl), EmoSet (Emotil and InSet) and SenSet (InSet and SentIl) with TFIDF. The result of this combination is used for machine learning classification with SVM and Logistic Regression algorithms. The model that has been built will be evaluated by comparing the accuracy value of cross validation. The largest and smallest accuracy results will be parameterised to improve the results of the average cross validation. The result of the greatest accuracy is SVM with a difference of 1.2% with Logistic Regression.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Analisis Sentimen, Figur Publik, Kamus Leksikon, Logistic Regression, Lexicon Dictionary, Sentiment Analysis, Public Figures, Support Vector Machine
Subjects:	T Technology > T Technology (General) > T57.5 Data Processing T Technology > T Technology (General) > T57.8 Nonlinear programming. Support vector machine. Wavelets. Hidden Markov models.
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Informatics Engineering > 55201-(S1) Undergraduate Thesis
Depositing User:	Fitriana Zahirah Tsabit
Date Deposited:	27 Oct 2023 04:31
Last Modified:	27 Oct 2023 04:31
URI:	http://repository.its.ac.id/id/eprint/100969

Actions (login required)

View Item