Sentiment Analysis for Udemy Reviews with Natural Language Processing and Machine Learning Methods

Authors : Sedat Sönmez, Tanju Açi, Hidayet Takcı, Hakan Kekül

Pages : 184-191

View : 103 | Download : 108

Publication Date : 2024-12-31

Article Type : Research Paper

Abstract :Understanding and classifying sentiment content in textual data is an important requirement for many industries. Text-based data such as social media platforms, customer feedback and product reviews are a rich source of human emotions and opinions. Extracting meaningful information from this text data and understanding the emotional content helps businesses make strategic decisions, develop products and improve their services. Machine learning methods are widely used to perform sentiment analysis on large amounts of text data. These methods are used to process text data, extract features, train models and classify emotional content. Natural language processing techniques are used to solve a range of problems such as increasing an application\\\'s user satisfaction, improving its services or optimising marketing strategies. In this study, emotional tones are determined by analysing course comments in the Udemy application. Prediction is made by classifying positive or negative comments. Udemy application comments on Google Play were used and sentiment analysis is performed using K-Nearest Neighbour (KNN) and Random Forest Classification (RFC) algorithms. As a result of the analyses, it was observed that the KNN algorithm predicted with 84% accuracy. Accuracy, F1 Score, Recall, Precision metrics were used as performance measures.
Keywords : Duygu analizi, Udemy, RFC Algoritması, KNN algoritması, makine öğrenmesi