Determining priority location for tourist village development using k-means clustering

• Main Authors: Ramadhani, Dinda Nurul, Rukmi, Hendang Setyo, Arif, Fahmi, Afifah, Alif Ulfa
• Format: Conference Proceeding
• Language: English
• Subjects: Cellular communication; Cluster analysis; Clustering; Corn; Data mining; Decision making; Farming; Industry standards; Local government; Tourism; Towns; Vector quantization; Villages
• ISSN: 0094-243X; 1551-7616
• DOI: 10.1063/5.0115121

Determination of an appropriate location is one of the most important strategic decisions in the development of tourist village. Unfortunately, numerous potential locations with various characteristics and attributes might increase the difficulty in finding the proper locations. The same problem happened in Sumedang, Indonesia, where the government has an intention to develop corn farming based tourist village. The data from Government's central bureau for statistics (BPS) showed that Sumedang as the corn producer has a lot of corn farming village with various characteristics. This study aimed to set priority for the corn farming based tourist village location in Sumedang using k-means clustering to help the government to conduct the process of decision making and development strategy for the appropriate location. This study adopted Cross Industry Standard Process for Data Mining (CRISP-DM) methodology that consists of six phases: business understanding, data understanding, data preparation, modeling, evaluation, and deployment. Using the data compiled from the database of local government agricultural office, tourism office and central bureau for statistics, this study defined the characteristics of the village through the attributes of the number of nearby banks, the number of BTS & number of cellular communication service operators, the number of minimarkets & shops, permanent & semi-permanent market shops, and data the number of small micro handicraft, food & beverage industries. As the result, with Davies Bouldin Index (DBI) of 0.103, this study grouped 277 alternative villages into 7 optimum clusters. Analysis of characteristics for each cluster was then conducted to find the priority level of each cluster. Based on its characteristics, cluster 5 that consists of 6 villages was selected as a location with the highest priority to be developed. This study showed that the use of k-means clustering could help in setting the priority group from numerous alternatives as it divided items into groups of similar characteristics.