Each student who attended the lecture at a college will inevitably undergo thesis examination or final project to finish the S1 degree. However, students still often have difficulty in determining the title theme to be lifted. On the other hand, many students take random theme of the thesis title or final project, just following friends or colleagues or looking for a single reference from the library. Therefore, it takes a process of data mining to assist students in determining the proper final assignment theme. The process of data mining is done by using K-Means algorithm with input value sudentscourses that have reached as the determining aspect, so that it found a pattern of students interest which is used to recommended themin determining the final project or thesis theme that suits their ability. In its application, classification as done against the final project themes were divided into 7 groups/clusters. Using students score as the input, the process is done using the K-Means algorithm by taking into account distance between data ti the center cluster (centroid). So that there is no more data move to the other cluster. Based on the testing against the centroid for 15 times, obtained a result that second centroid has the highest truth compared to other centroid with the value was 90.24%. Therefore, the system will use the second centroid as a new reference in determining the theme of the final project for each student by viewing the closest data to the centroid point.
展开▼
