Thesis:

Classification of customer reviews using improved decision tree and Tabu search algorithm

"Data classification means dividing data into different groups based on common characteristics and is of great importance in data mining. This technique can be effective in predicting events, making better decisions, identifying patterns, and improving customer experiences. Metaheuristic algorithms, including the tabu search algorithm, are designed to solve complex optimization problems and can be applied in classifying customer reviews in the hotel .industry. By using these algorithms, companies can analyze customer feedback and take appropriate actions to improve services and increase customer satisfaction The proposed method in this research involves classifying customer reviews using an improved decision tree and the tabu search algorithm. In this study, the TripAdvisor.com database was examined, and after preprocessing and feature extraction through feature selection using the forbidden search algorithm, we classified customer reviews using a decision tree. In this research, for the values of Accuracy, Recall, Precision, and F-score with 10 features, the results were 99.6%, 99.7%, 99.6%, and 99.6%, respectively. The results indicate that the proposed method performed better than other studies reviewed.

Keywords: Data mining, Customer Reviews, Classification, Improved Decision Tree, Tabu Search Algorithm