Sentiment Analysis on Official News Accounts of Twitter Media in Predicting Facebook Stock

Jessica, Jessica and Oetama, Raymond Sunardi (2020) Sentiment Analysis on Official News Accounts of Twitter Media in Predicting Facebook Stock. In: 5th International Conference on New Media Studies, 9-11 Oct. 2019, Bali, Indonesia.

Full text not available from this repository.

Abstract

Social media like Twitter is a place for people who are interested to talk about financial news faster, so that it leads to interesting insights. Twitter also generates a lot of data that may relate to people's sentiments about a company. These factors can be considered for stock predictions. A successful stock market prediction is not only for achieving a high level of results, but also minimizing inaccuracies in prediction. This study aims to perform some experiments using some trading methods that help to increase the effectiveness of the stock price predictions of Facebook Company. Sentiment analysis is applied on a collection of tweets that taken from several official news offices on Twitter and combined with some moving average methods of 5, 10 and 15 days. The experimental results show that the prediction method with moving averages of 5 days without sentiments reaches the most profit. However, by combining this method with sentiment analysis on CNBC news data it results more effective shown on its average daily profit.

Item Type: Conference or Workshop Item (Paper)
Keywords: Text Mining , Sentiment Analysis , Moving Average , Stock Prediction
Subjects: 000 Computer Science, Information and General Works > 000 Computer Science, Knowledge and Systems > 004 Computer Science, Data Processing, Hardware
000 Computer Science, Information and General Works > 000 Computer Science, Knowledge and Systems > 006 Special Computer Methods
Divisions: Faculty of Engineering & Informatics > Information System
Depositing User: Administrator UMN Library
Date Deposited: 25 Nov 2021 09:32
Last Modified: 13 Apr 2023 03:13
URI: https://kc.umn.ac.id/id/eprint/19295

Actions (login required)

View Item View Item