Turkish-Text-Classification

In this project, sentiment analysis is performed with Getir comments.

Dataset

The dataset contains 74,500 labeled data collected from the Play Store for the Getir application. Of these data, 54,000 were labeled as positive (1) and 20,000 as negative (0). You can refer to this file for data collection preprocessing steps.Bert was used for data labeling.
You can refer to this document for data labeling steps.

Machine Learning Algorithms

In the project, SVM, Naive Bayes, MLP and RNN machine learning algorithms were used for sentiment analysis.