Text Mining Implementation of Text Mining to Know Someone's Social Pattern Based on Tweet on Twitter

Implementation of Text Mining to Know Someone's Social Pattern Based on Tweet on Twitter

  • martin adrian nanda Politeknik Caltex Riau

Abstract

Twitter social networks are information technologies that allow users to access the information they disseminate. Twitter activities in Indonesia use many of Twitter's features to present all forms of information that he wants to write. Verbal communication on Tweet may be one of those who identify social patterns. Currently, human social models are necessary for personal or group interests. Information about social models can also be used to make effective and efficient decisions without the help of a psychologist. Then it takes time and special knowledge to reveal a lot of tweets. To cope with this, a system will be created that can identify social models of a person based on tweet on Twitter. Tweet is acquired and processed using text mining. Then, in the classification process, the naive Bayes classifier method will be used. Text mining methods and the naive Bayes classifier method have been successfully applied in defining social patterns in the system. There are 4 tests, such as: K-fold Cross Validation testing, Blackbox Testing, Questionnaire, and Psychologist Validity. From these tests it can be concluded that the system has been tested for functionality, and obtained an accuracy of 96% from the K-fold Cross Validation and 90% from the validity of the psychologist.   Keywords: Information Technology, Social Media, Twitter, Tweet, Text Mining, Naive Bayes Classifier
Published
2020-09-10
Section
Artikel