ANALISIS SENTIMEN MASYARAKAT TERHADAP TRADING BINARY OPTION PADA MEDIA SOSIAL TWITTER MENGGUNAKAN KLASIFIKASI SUPPORT VECTOR MACHINE

ADITYA, DWI WAHYU (2023) ANALISIS SENTIMEN MASYARAKAT TERHADAP TRADING BINARY OPTION PADA MEDIA SOSIAL TWITTER MENGGUNAKAN KLASIFIKASI SUPPORT VECTOR MACHINE. S1 thesis, Universitas Muhammadiyah Purwokerto.

Preview	Text DWI WAHYU ADITYA_COVER.pdf Download (34MB) | Preview
Preview	Text DWI WAHYU ADITYA_BAB I.pdf Download (1MB) | Preview
Preview	Text DWI WAHYU ADITYA_BAB II.pdf Download (1MB) | Preview
	Text DWI WAHYU ADITYA_BAB III.pdf Restricted to Registered users only Download (1MB)
	Text DWI WAHYU ADITYA_BAB IV.pdf Restricted to Registered users only Download (1MB)
	Text DWI WAHYU ADITYA_BAB V.pdf Restricted to Registered users only Download (1MB)
Preview	Text DWI WAHYU ADITYA_DAFTAR PUSTAKA.pdf Download (1MB) | Preview
	Text DWI WAHYU ADITYA_LAMPIRAN.pdf Restricted to Registered users only Download (6MB)

Abstract

People talk about hot topics through social media. One of the most widely used social media is Twitter. There are also various topics discussed on Twitter, such as binary options trading which has recently been widely discussed by Indonesian people and had become a trending topic on Twitter. Binary options trading is an instrument that is still relatively new to the people of Indonesia. The public only became aware of the existence of binary options after there was news about several option holders who were tricked into feeling and experiencing big losses as a result of joining the binary options platform. New binary options platform users still do not realize that high risk will always be associated with the projected price movement of the financial product used as the underlying asset. In this study, data was obtained by scraping from Twitter using the snscrape library. Data from scraping results obtained for 16,084 tweets. From this data preprocessing is carried out with the first stage of case folding, then cleaning, tokenizing, stopward removal, word normalization and finally stemmiong. After the data is preprocessed, the remaining 8,017 data are used for the next process, namely data labeling. The data labeling process was carried out using the lexicon senticnet 7 method and obtained a total of 3,412 data of positive sentiment, 3,780 data of negative sentiment and 825 data of neutral sentiment. Then a support vector machine classification algorithm is applied using a linear kernel with a training and test data ratio of 9:1. The results of tests conducted on tweet trading binary options produce an accuracy value of 90.6%, a precision of 88.7%, a recall of 92.4%, and an f1-score of 90.3%

Actions (login required)

View Item