Classification of Turkish E-Commerce Product Reviews

Nowadays, e-commerce product reviews play a very important role in online shopping. With the importance of technology in our lives and the intense interest in online shopping, the classification of these comments with text classification is quite important. In addition to supervised and unsupervised machine learning algorithms, which are very useful and effective in text classification, deep learning algorithms are also very popular and have high success rates. The study aims to provide a brief overview of machine learning methods for text classification. In the dataset used, the comments under a product on online shopping sites were collected and a dataset was created by giving class labels according to whether the comments were positive, negative, or neutral. The dataset has 6799 positive, 6978 negatives, and 1393 neutral comments, including 15170 comments. Convolutional Neural Network (CNN) is proposed as a classifier in the classification phase. Also, the proposed method is compared with Long Short Term Memory (LSTM), Bidirectional Long Short Term Memory (Bi-LSTM), and Support Vector Machine (SVM).