The main goal of this project is to create an application able to filter and classify a big amount of messages (such as tweets or e-mails) according to the value of their fields. By doing this, the user can discard those which are irrelevant and focus on the important ones and, finally, establish connections between them . In order to achieve this, the user can tag messages and individuals (who are the senders and/or receivers of the messages) manually, but also in an automatic way using Natural Language Processing algorithms to detect similar messages. The application will be developed in Java, using Java Swing for the graphic user interface and a SQLite database for saving the generated data. Additionally, the application will work as a framework for adding more techniques of detection of similar messages, in order to compare the results of each technique and select the most appropriate one for the kind of message we are working with.