Sentiment Analysis Dataset and Web Application for Turkish Tweets

Abstract

Today, Twitter (X ) is one of the most essential and popular social networking sites. It is very important to analyze the sentiments of the tweets posted on this platform to understand people and understand their opinions on any topic. Thus, you can determine what people are thinking and talking about on a topic you can choose, such as a brand, business, economy, or education. In this study, a dataset is created with Turkish tweets collected using the Twitter API. Then, techniques such as Word2Vec and Bag of Words (BoW) are used to clean this dataset and use it more comfortably. Afterward, this cleaned dataset is classified as Positive, Negative, and Neutral using classification methods such as Decision Tree, Logistic Regression, Support Vector Machine (SVM), Random Forest, and XGBClassifier. Finally, a simple website has been created using JavaScript for users to use this application efficiently.