| Summary: | The aim of the research is to compare the sentiment analysis methods used to reveal and classify the emotional tendencies in Turkish and English comments of hotel users. Within purpose, classification algorithms such as Decision Tree and Random Forest from machine learning methods were used. The data was obtained from Tripadvisor tourism portal with web scraping/mining technique within the scope of this study, which shows quantitative research feature. A purposeful sampling method was used in this study. Emotion analysis, which is one of the text mining applications, was used to analyze the data. KNIME Analytics Platform was used in the data analysis process. As a result of the research, it was seen that the machine learning algorithms performed more effective classification than dictionary-based analysis. In addition, the machine learning algorithms produced more successful results in the Turkish language comments at the classification stage.
|
|---|