Decision Support System to Determine Promotional Methods and Targets with K-Means Clustering

Promotion becomes one of the important aspects of institutions of college. The number of competitors demanding the marketing must be fast and accurate in formulating strategies and decision making. Data warehouse and data mining become one of the means to build a decision support system that can provide knowledge and wisdom quickly to be taken into consideration in promotion strategy planning. Development of this system then does the process of testing with the number of data 6171 rows of student enrollment taken directly from a transactional database. The data is done ETL process and clustering with the k-means clustering algorithm, then the data in each cluster is done grouping and summarization to get weighting. After that just done ranking to produce wisdom, one of them determine the list of schools that will be the target roadshow. The analysis also produces several patterns of student enrollment, namely the registrant pattern from the wave of registration and favorite or non-favorite school categories. In addition, the results of system design in this study can be developed easily if requires added external data. Such as data of SMK/SMK school graduates in the area or data of students enrolling in other universities. This is one of the superiority of semantic-based data warehouses.