Skip to content

guilherme-vaz/classificador-tcc

Repository files navigation

Ferramenta para classificação de sentimentos em tweets online

A ferramenta tem a capacidade de buscar tweets que possuam um termo específico no corpo de seu texto para então analisá-los e aferir um sentimento ao tweet. Os experimentos executados possibilitaram demonstrar que a ferramenta proposta é capaz de obter resultados consistentes e satisfatórios para uma vasta diversidade de termos. Os resultados demonstraram uma acurácia de 84,6% em testes com base de dados estática, e uma taxa de acertos de 80% em um experimento com base de dados online.

Esse projeto é fundamentado em:

  • Análise de Sentimentos
  • Mineração de Texto
  • Processamento de linguagem natural
  • Aprendizado de máquina supervisionado

An online tool for rating feelings on tweets

The tool has the ability to search tweets that have a specific term in its text body, it then analyze the tweet and give it an emotion (positive or negative). The experiments carried out allowed to demonstrate that the proposed tool is capable of get consistent and satisfying results for a wide variety of terms. The results show an accuracy of 84,6% in tests with a static database, and a 80% hit rate in an experiment with an online database.

This project is based on:

  • Sentiment analysis
  • Text mining
  • Natural language processing (NLP)
  • Supervised Machine Learning

🛠 Tecnologias // Tools

As seguintes ferramentas foram usadas na construção do projeto:

Run project

python manage.py runserver