<![CDATA[At present, with the ease of access to information, many tourist sites use rating features to help facilitate information. Rating is used as an indicator to support quality and popularity. Users only give an overall assessment of each comment and do not provide an assessment in accordance with the aspects discussed, making it difficult for comment readers to analyze the superior aspects of the comment. From this problem, in this study a rating classification system will be made on tourist attractions using the Fuzzy K-Nearest Neighbor (FKNN) method. FKNN method is one of the development methods of the KNN method, the difference is that there is a membership class to determine the classification class. In addition, this study uses a Lexicon Based dictionary to determine feature extraction. The results of the tests in this study showed the highest accuracy of K=20 values of 60% while the accuracy of precision and recall values reached 40% and 40% respectively. In testing the K-Fold Cross Validation with 5 fold it produces an average of 51.4%.]]>
Copyrights © 2019
