<![CDATA[The students' thought process during learning reflects their understanding and misunderstanding about the structure of the questions in an exercise. The application of understanding of the structure of questions has been packaged in a learning media called Monsakun. Monsakun provides support by offering the concept of understanding the structure of questions in solving simple arithmetic word problems (addition and subtraction). Although Monsakun has successfully provided support in learning, identification of group learning patterns among students has not been done. This pattern grouping needs to be done to make it easy for teachers to understand the characteristics of students' thinking, understand the difficulties faced, and provide feedback in accordance with the characteristics of thinking and difficulties experienced by these students. This study aims to group students' thought processes while studying at Monsakun using the K-Means algorithm that is optimized with the Particle Swarm Optimization algorithm in determining initial centroids. The data used is a Monsakun level 5 datalog consisting of 12 questions. Based on the implementation and testing that has been done, the results of grouping are dominated by 2 clusters where the quality of the cluster is determined using the Silhouette Coefficient method.]]>
Copyrights © 2020
