Representasi Word Graph Frasa Preposisional Bahasa Indonesia Menggunakan XML

Abstract

Knowledge Graph (KG) is one of the methods in natural language processing that can be used to represent either knowledge or information in a form of a graph. KG is also used to minimize the problem of ambiguity in text processing. KG has been applied to form word graphs for prepositions. The objective of this research is to represent word graphs for Indonesia prepositional phrases using XML. We developed a Java based system to accept a prepositional phrase and determine the appropiate pattern for it. The system would then generate an XML document to represent the pattern. This research uses 51 preposition words and 12 types of accompanying word preposition. The total of accompanying word prepositions are 649 words. There are 127 kinds of prepositional phrase patterns formed in this research and 26 patterns of word graph prepositions represented in an XML document. The root for each XML word graph document preposition is called Graph Element. The Graph Element comprises of two leaves elements: Components and Relationships. The Components contains the list of elements used in forming the pattern of word graph preposition. The Relationships contains the relation between elements used in the Components.