| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 1.54 MB | Adobe PDF |
Abstract(s)
Numa sociedade cada vez mais ligada pelas redes sociais é importante saber ao certo o que aí se discute, tanto sobre marcas específicas ou produtos, como sobre eventos importantes que ocorram. Esse conhecimento é de extrema importância para as entidades responsáveis pelos produtos ou eventos em questão que querem melhorar a sua prestação com o intuito de agradar cada vez mais aos utilizadores. Isto pode ser feito de uma maneira muito fácil através da monitorização de redes sociais que, como foi dito, são um meio de comunicação em crescimento constante.
A presente dissertação insere-se no contexto da análise de sentimentos baseada na extração de informação das redes sociais. Com o objetivo de se validar o modelo proposto, dois casos de estudo foram considerados: análise da competição Mundial FIFA 2018 de futebol e o “efeito Cristiano Ronaldo”.
Esta dissertação pretende fazer o estudo de conteúdos da rede social Twitter, não se restringindo apenas ao que é escrito, mas focando-se também nas emoções que são expressas nessas publicações.
São referidas nesta dissertação, relativamente ao primeiro caso de estudo, as quatro seleções que chegaram mais longe no campeonato do mundo de futebol 2018, que foram: França, Croácia, Bélgica e Inglaterra. Além destas seleções é mencionada a de Portugal, uma vez que foi a vencedora da competição UEFA Euro 2016. Por fim, é feito o levantamento das emoções expressas pelos utilizadores em relação a cada seleção em particular, além do levantamento das emoções provocadas em cada jogo entre estas seleções.
Relativamente ao segundo caso de estudo, é referida a transferência do jogador Cristiano Ronaldo e ainda os clubes Real Madrid CF e Juventus FC que são o seu antigo e atual clubes, respetivamente.
Apesar de já existirem bastantes trabalhos realizados nesta área, tal não acontece concretamente sobre os dois casos de estudo propostos, o que torna mais interessante o tema desta dissertação.
No final da realização desta dissertação, relativamente ao primeiro caso de estudo, podemos concluir que a seleção que provoca sentimentos mais positivos, dentro do conjunto de todas as seleções, é a seleção francesa. Isto devido à vitória da mesma no mundial de futebol de 2018. Ainda no primeiro caso de estudo relativamente à identificação das emoções demonstradas, podemos concluir que as emoções predominantes são a de “felicidade”, seguida pela emoção de “surpresa”. Esta última pode ser explicada pela chegada da seleção croata à final da competição, que não fazia parte do grupo das seleções teoricamente favoritas.
Relativamente ao segundo caso de estudo, na identificação de emoções, com o objetivo de perceber como é que a transferência do jogador Cristiano Ronaldo afetou os adeptos afetos às duas equipas, concluímos que, relativamente aos adeptos da Juventus, foi predominante o sentimento de “surpresa”, porque esta transferência não parecia ser possível. Relativamente aos adeptos do Real Madrid, a emoção predominante foi também de “surpresa”, pelas mesmas razões dos adeptos da Juventus. No entanto, foi identificado um número significante de tweets que expressaram “tristeza” e “medo”. Isto devido ao jogador ser considerado por muitos, o melhor jogador do mundo.
Adicionalmente nesta dissertação, é apresentado um modelo constituído por etapas, que propõe uma diretriz para a realização deste tipo de projetos, relativos à Mineração de Texto e Análise de Sentimentos.
In a society increasingly linked by social networks, it is important to know for sure what is said there, both on specific brands or products, and on important events that occur. This knowledge is extremely important for the entities responsible for the products or events in question that want to improve their performance in order to please the users more and more. This can be done very easily by monitoring social networks which, as has been said, are a constantly growing medium of communication. The present dissertation is inserted in the context of the sentiment analysis based on the extraction of information from social networks. In order to validate the proposed model, two case studies were considered: analysis of the FIFA 2018 World Cup competition and the Cristiano Ronaldo move from Real Madrid CF to Juventus FC. This dissertation intends to study the contents of the social network Twitter, not only being restricted to what is written, but also focusing attention on the emotions that are expressed in those publications. In this dissertation, in relation to the first case study, the four teams that went the furthest in the competition were: France, Croatia, Belgium and England. In addition to these selections Portugal is mentioned since it was the winner of the UEFA Euro 2016 competition. Finally, it is made the survey of the emotions expressed by the users in relation to each particular selection, as well as the emotions raised in each game between these selections. Regarding the second case study, the Cristiano Ronaldo player is mentioned, as well as the clubs Real Madrid CF and Juventus FC which are his old and current clubs, respectively. Although there are already many researches done in this area, this does not happen concretely on the two proposed case studies, which makes the topic of this dissertation more interesting. At the end of this dissertation, in relation to the first case study, we can conclude that the the national team that provokes more positive sentiment, within the set of all national teams, is the French national team. This is due to the victory in the FIFA World Cup 2018. Still in the first case study, regarding the emotion detection, we can conclude that the predominant emotions are the one of “happiness”, followed by the emotion of “surprise”. This last can be explained by the presence of the Croatian national team in the final, that was not part of the group of the national teams theoretically favorites. Regarding the second case study, in the emotion detection, in order to understand how the move of the player Cristiano Ronaldo affected the fans of both teams, we concluded that, relatively, the predominant emotion was also “surprise”, for the same reason as the Juventus fans. However, a significant number of tweets expressing “sadness” and “fear” were identified. This is due to the player being considered by many, the best player in the world. Additionally, in this dissertation, is presented a model consisting of stages, which proposes a guideline for the accomplishment of this type of projects, related to Text Mining and Sentiment Analysis.
In a society increasingly linked by social networks, it is important to know for sure what is said there, both on specific brands or products, and on important events that occur. This knowledge is extremely important for the entities responsible for the products or events in question that want to improve their performance in order to please the users more and more. This can be done very easily by monitoring social networks which, as has been said, are a constantly growing medium of communication. The present dissertation is inserted in the context of the sentiment analysis based on the extraction of information from social networks. In order to validate the proposed model, two case studies were considered: analysis of the FIFA 2018 World Cup competition and the Cristiano Ronaldo move from Real Madrid CF to Juventus FC. This dissertation intends to study the contents of the social network Twitter, not only being restricted to what is written, but also focusing attention on the emotions that are expressed in those publications. In this dissertation, in relation to the first case study, the four teams that went the furthest in the competition were: France, Croatia, Belgium and England. In addition to these selections Portugal is mentioned since it was the winner of the UEFA Euro 2016 competition. Finally, it is made the survey of the emotions expressed by the users in relation to each particular selection, as well as the emotions raised in each game between these selections. Regarding the second case study, the Cristiano Ronaldo player is mentioned, as well as the clubs Real Madrid CF and Juventus FC which are his old and current clubs, respectively. Although there are already many researches done in this area, this does not happen concretely on the two proposed case studies, which makes the topic of this dissertation more interesting. At the end of this dissertation, in relation to the first case study, we can conclude that the the national team that provokes more positive sentiment, within the set of all national teams, is the French national team. This is due to the victory in the FIFA World Cup 2018. Still in the first case study, regarding the emotion detection, we can conclude that the predominant emotions are the one of “happiness”, followed by the emotion of “surprise”. This last can be explained by the presence of the Croatian national team in the final, that was not part of the group of the national teams theoretically favorites. Regarding the second case study, in the emotion detection, in order to understand how the move of the player Cristiano Ronaldo affected the fans of both teams, we concluded that, relatively, the predominant emotion was also “surprise”, for the same reason as the Juventus fans. However, a significant number of tweets expressing “sadness” and “fear” were identified. This is due to the player being considered by many, the best player in the world. Additionally, in this dissertation, is presented a model consisting of stages, which proposes a guideline for the accomplishment of this type of projects, related to Text Mining and Sentiment Analysis.
Description
Keywords
Cristiano Ronaldo Juventus Fc Mineração de Dados Mineração de Texto Mundial Fifa 2018 Processamento de Linguagem Natural Real Madrid Cf Twitter