CLUSTERING K-MEANS JENIS KATA PADA LAPORAN KEGIATAN KULIAH KERJA NYATA (KKN) UNIVERSITAS LAMPUNG MENGGUNAKAN WORD2VEC

Abstract

Kuliah Kerja Nyata (KKN) is a form of student service activities for the community, requesting and developing science and technology carried out off-campus within a period, linking work, and special requirements managed by the Badan Pelaksana Kuliah Kerja Nyata (BP-KKN). While carrying out KKN activities, each group of students is required to upload a report of the activities carried out in the village. In uploading the report file, there are several categories in each activity, including socialization, training, and character development. To classify the results of uploading activities one of which can be done using clustering techniques. In this research, a clustering of discussion on KKN student activities will be conducted at the University of Lampung. The text mining method is used to process KKN student activities to be more structured. Information on the KKN student activities was obtained as a feature with the Word2Vec weighting technique. The algorithm used is the K-Mean algorithm which has a high accuracy of the size of the object, so this algorithm is relatively more measurable and efficient for processing large numbers of objects. From the results of research conducted, it has been found that apply the text mining process algorithm for clustering with the K-means method on the Unila KKN Student activity data produces a value of k = 2, a lot of filtered data in the preprocess is 6284 data, using this method has not yet gotten a good association analysis because the results of the second cluster do not show the general types of words, typos and reporting activities by students who are not specifically can affect the results of clustering that is not good.