<![CDATA[Smartlegal.id is a company engaged in the field of law. As of 2021, SmartLegal.id has taken care of more than 60,492 business legalities throughout Indonesia since 2014. According to Smartlegal.id's Product Specialist Manager, analysis of client behavior with the aim of forming marketing strategies is still experiencing time constraints, because it requires scheduling face-to-face meetings with clients. In addition, the marketing problem that often occurs is the alignment between content, events, promos and also market needs which can take a long time to match and align. Based on these problems, a practical analysis of client behavior is needed to save time. One technique that can be used is clustering. The K-Means method is a method that has been widely implemented by Information Technology practitioners. However, with the high dimensions of the existing data, it is necessary to adjust the method by adding a kernel function so that it can better classify non-linearly separable data. From the results of the research conducted, the best Silhouette Score was 0,9035 using 2 clusters, the Polynomial kernel function with the Polynomial degree parameter was 30, and the data percentage was 100%. This study also conducted a comparison of the effectiveness between K-Means and Kernel K-Means in segmenting.]]>
Copyrights © 2022
