Open AccessNaïve Bayes Classifiers For Tweet Sentiment Analysis Using GPU
Author(s) -
Islamiyah,
Nataniel Dengen,
Eny Maria
Publication year - 2019
Publication title -
international journal of engineering and advanced technology
Language(s) - English
Resource type - Journals
ISSN - 2249-8958
DOI - 10.35940/ijeat.e1216.0585c19
Subject(s) - computer science , naive bayes classifier , preprocessor , weighting , the internet , data pre processing , sentiment analysis , construct (python library) , data mining , process (computing) , artificial intelligence , information retrieval , machine learning , world wide web , support vector machine , medicine , radiology , programming language , operating system
The use of computers to solve problems has been done for all areas of work. Along with this, demanded faster computing process. To perform sentiment analysis of data obtained from the internet. Data taken from micro-blogging which at this time became the most popular communication tool and favored by internet users. The method used to construct the classification model of training data in this research is Naive Bayes Method. Training data is collected by utilizing the crontab facility with query emoticons and national media accounts linked to the Twitter API. The collected data will pass certain preprocessing before the training. The weighting feature used is the term frequency with TF-IDF. All data used in this research is a tweet that is delivered in Bahasa Indonesia. From the implementation results obtained 96.61% accuracy for sequential classification conducted using GPU GeForce 930M.