Analisa Extrasi Informasi Pada Abstraksi Jurnal Skripsi Berbahasa Indonesia Menggunakan Algoritma K Nearest Neighbor

Abstract

Information Extraction is the extraction of structured facts and information from the contents of a large collection of texts. The definition of facts here is a variety of entities that are calculated or connected in the form of structured information as input to the database. Processing information extraction data in thesis journal abstracts using the KNN algorithm starts from the data selection stage (attributes used and determination of training data and data testing), the algorithm testing stage (KNN), and the accuracy test stage (using split validation). The classification process in thesis journal abstracts using the KNN algorithm is one way to classify information extraction in thesis journal abstracts. The classification process in the thesis journal abstract using the KNN algorithm is used to avoid information extraction errors in the thesis journal abstract. processing data starting from the data preprocessing stage and text mining calculations consisting of weighting term frequency and weighting concept frequency and Cosine Similarity D7 0.0332, D15 0, D10 0.1296, D14 0.1296

Keywords: Text Mining, Information Extraction, K-NN