Visualisasi dan Navigasi Semantik pada Subjek Tesis IPB Menggunakan Peluang Bersyarat

Abstract

Searching and browsing research documents, such as thesis, using subject terms in document is an effective way to obtain information about existing research in IPB. These two tasks results will be more interactive when a visualization module is added. This research aims to build a document browsing system based on subjects of the IPB’s thesis metadata and to visualize the relations among subject terms. Subject terms are stemmed using the Porter Stemmer algorithm and the list of subject term are created using the thresholding technique. Analysis on threshold values is performed to determine the threshold level which gives efficient computation time and the optimal number of subject terms. The experimental results show that the threshold value of 4 provides the computation time of 334 seconds and results in 381 subject terms. Furthermore, the probabilities of subject terms are calculated using the conditional probability formula and a visualization module is designed.