Banca de QUALIFICAÇÃO: GILSON DA SILVA VASCONCELOS

Uma banca de QUALIFICAÇÃO de MESTRADO foi cadastrada pelo programa.
DISCENTE : GILSON DA SILVA VASCONCELOS
DATA : 26/09/2023
HORA: 16:00
LOCAL: Video Conferência
TÍTULO:

THE INFLUENCE OF TWEETS SENTIMENT TO PREDICT IBOVESPA STOCKS


PALAVRAS-CHAVES:

Sentiment Index; IBOVESPA forecast; Technical Indicators; Twitter


PÁGINAS: 28
GRANDE ÁREA: Ciências Sociais Aplicadas
ÁREA: Economia
RESUMO:

The research will analyze the influence that twitter data has to predict the shares that make up the IBOVESPA, as it is noticeable that the use of textual data has been widely used to identify the effect of investor sentiment regarding the prediction of stock returns. This textual information is available through various means of communication such as Twitter, through tweets that can be extracted through its API. Thus, sentiment index will be used. Medeiros et.al (2023) in order to verify the texts that mention the IBOVESPA. From this, it is intended to compare, through the behavior of the indices, the performance of the sentiment of twitter, with the Technical Analysis and with the historical index of the Brazilian stock market. Furthermore, we intend to calculate the utility gain of an individual investor for each model. For this, data will be used with daily frequency starting on January 1, 2007 to April 1, 2022, which have 1830 observations. Then, let's analyze whether textual sentiment indices can outperform more macroeconomic models, as well as models with time-varying words. So we can find the best predictor from the textual data using machine learning over time and create a sentiment index through the tone of the words


MEMBROS DA BANCA:
Presidente - 056.549.504-60 - LUCAS LÚCIO GODEIRO - UFERSA
Interna - 3295 - MARIA ELZA DE ANDRADE
Interno - 720.525.114-15 - LAURO CÉSAR BEZERRA NOGUEIRA - UFERSA
Notícia cadastrada em: 14/09/2023 09:14
SIGAA | Superintendência de Tecnologia da Informação - STI/UERN - (84) 3315-2222 | Copyright © 2006-2024 - UFRN - app01-uern.info.ufrn.br.app01-uern