| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 6.67 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
Desde que existe, o Homem sentiu a necessidade de deixar a sua marca no mundo. Com a invenção da escrita, este passou a poder registar os seus pensamentos e ideias em algo mais duradouro do que a sua memória.
Com o evoluir do mundo, foram aparecendo novas formas de poder deixar registos escritos.
Com o aparecimento de novas tecnologias e a aproximação das pessoas online, também a forma de registar os pensamentos mudou. Qualquer pessoa pode escrever e partilhar online, para que outras possam ler.
Neste momento já existem várias plataformas, cujo objetivo é facilitar a partilha de histórias criadas por pessoas que desejam partilhá-las com o mundo. No entanto, muitos destes escritores são amadores e por isso podem não ter os conhecimentos necessários para conseguir caracterizar as suas histórias adequadamente.
Estas plataformas oferecem funcionalidades de pesquisa e recomendação rudimentares.
A aplicação de técnicas de data mining e text mining podem ajudar os escritores a caracterizar as suas histórias de uma forma automática, eficiente e sem exigir conhecimentos literários.
Com esta automação podemos obter dados que podem ser também utilizados para recomendar histórias e disponibilizar funcionalidades de pesquisa e recomendação, por exemplo, por personagem, género literário ou ainda por outras características.
O objetivo desta dissertação é implementar estas tecnologias numa plataforma para escrita, partilha e leitura de histórias dos mais variados géneros literários, por parte de escritores amadores e seus fãs, para que a sua vida seja facilitada, quer ao nível da escrita, com o preenchimento automático de atributos descritores nas histórias — como o género literário ou a lista de personagens— quer ao nível da sugestão de conteúdos aos utilizadores.
Since Man exists, he felt the need to leave his mark on the world. With the invention of writing, Man was able to record his thoughts and ideas into something more lasting than memory. As the world evolved, new ways of leaving written records began to appear. With the emergence of new technologies and the approximation of people online, the way of recording thoughts also changed. Anyone can write and share online, so that others can read. At the moment there are already multiple platforms, whose goal is to facilitate the sharing of stories created by people who want to share them with the world. However, many of these writers are amateurs and they may not have the necessary knowledge to be able to characterize their stories appropriately. These platforms offer rudimentary search and recommendation features. The application of data mining and text mining can help writers to characterize their stories automatically, efficiently and without requiring literary knowledge. With this automation we can obtain data that can also be used to recommend stories and search and recommendation features like, character, literary genre or other characteristics. The goal of this dissertation is to implement these technologies on a platform for writing, sharing and reading stories of various literary genres, by amateur writers and their fans, so hat their life is facilitated, both in terms of writing, with the auto-completion of attribute descriptors in the stories—like the literary genre or the list of characters—and in terms of the suggestion of content to users.
Since Man exists, he felt the need to leave his mark on the world. With the invention of writing, Man was able to record his thoughts and ideas into something more lasting than memory. As the world evolved, new ways of leaving written records began to appear. With the emergence of new technologies and the approximation of people online, the way of recording thoughts also changed. Anyone can write and share online, so that others can read. At the moment there are already multiple platforms, whose goal is to facilitate the sharing of stories created by people who want to share them with the world. However, many of these writers are amateurs and they may not have the necessary knowledge to be able to characterize their stories appropriately. These platforms offer rudimentary search and recommendation features. The application of data mining and text mining can help writers to characterize their stories automatically, efficiently and without requiring literary knowledge. With this automation we can obtain data that can also be used to recommend stories and search and recommendation features like, character, literary genre or other characteristics. The goal of this dissertation is to implement these technologies on a platform for writing, sharing and reading stories of various literary genres, by amateur writers and their fans, so hat their life is facilitated, both in terms of writing, with the auto-completion of attribute descriptors in the stories—like the literary genre or the list of characters—and in terms of the suggestion of content to users.
Description
Keywords
Machine learning Natural language processing (NLP) Inteligência artificial Processamento de texto Text Mining Data Mining Artificial intelligence Text processing