Natural Language Processing in Meaning Representation for Sentiment Analysis in Serbian Language
Marko M. Živanović, Olga Ristić, Sandra Milunović Koprivica · 2024
This paper explores machine learning algorithms that contribute to meaning representation and context modeling in sentiment analysis. Language preprocessing techniques are described in detail. The study also discusses string distance calculations and the application of Naive Bayes for classification, emphasizing important model metrics such as accuracy. The final section of the paper presents a practical example encompassing the process of data collection, analysis, preprocessing, classification using machine learning algorithms, and model evaluation. Testing demonstrated the system’s ability to classify sentiments in Serbian Language.