<![CDATA[Simpang Lima Gumul (SLG) is a monument that has become an iconic building as well as a tourist destination in Kediri. Visitors who come can provide reviews on Google Review SLG to help the manager know the advantages and disadvantages of existing infrastructure. However, the SLG manager does not have a system that can automatically classify positive and negative reviews. This problem can be solved by using a sentiment analysis system. The sentiment analysis system used in this study uses Neighbor-Weighted K-Nearest Neighbor (NWKNN) and BM25 methods. The stages of this system include preprocessing process, weighting TF-IDF, ranking using BM25, and classification process using NWKNN. The number of data used is 1000 data, with the division of 800 training data and 200 test data. The test was carried out using 5-fold cross validation to test the k and exponential values in the NWKNN method and the k1 and b values ​​in the BM25 method. Based on the tests carried out on each tested parameter, it was found that the best value for the parameter value k1 = 1.2, b = 0.5, k = 20, and exponent = 2. The combination of these parameter values ​​produces an average value precision of 0.9509, recall of 0.9589, accuracy of 0.93, and f-measure of 0.9548.]]>
Copyrights © 2021
