<![CDATA[The challenge in maintaining the business continuity of hotel, homestay, and resort accommodation services is the use of technology and marketing methods that are effective and efficient. This study aims to apply the Cross Industry Standard Process for Data Mining (CRISP-DM) method to optimize hotel, homestay, and resort marketing strategies. This research uses case studies on hotel, homestay, and resort accommodation services based on the context of the service provider's business operational area in North Toraja Regency, South Sulawesi Province. The results of this study show that TF-IDF results and sentiment extracts can be utilized for segmenting and targeting processes. Based on the results of TF-IDF, it is known that the main emphasis of hotel, homestay, or resort guests (Toraja Herigate, Toraja Misiliana, Rosaliana Homestay, Hotel Pison, and Luta Resort Toraja) with business operational areas in Rantepao sub-district, North Toraja Regency, South Sulawesi Province, is as follows: staff (97), Rantepao (98), clean (108), breakfast (139), nice (147), rooms (165), toraja (177), good (189), room (268), hotel (331). In addition, modeling results using the SVM algorithm with SMOTE operators can be used in positioning strategies to improve the competitiveness of accommodation services based on business operational areas. Specifically, the performance of the SVM algorithm using SMOTE operators resulted in an accuracy value of 97.64%, a precision value of 100%, a recall value of 95.37%, an f-measure value of 97.56%, an AUC value of 100%). Compared to the DT algorithm, the performance of SVM algorithms using SMOTE operators has better performance in the classification of review data. Thus, the CRISP-DM method is able to optimize the marketing strategy of hotels, homestays, and resorts.]]>
Copyrights © 2023
