Ontology learning framework for Quran

Ontology is an important element in the Semantic Web. It can represent scattered knowledge from unstructured text into a structured representation. However, extracting conceptual knowledge for the constructing of ontology is a time consuming task. Thus, the task is usually performed either automatic...

Full description

Bibliographic Details
Published in:Advanced Science Letters
Main Author: Ismail R.; Rahman N.A.; Bakar Z.A.
Format: Article
Language:English
Published: American Scientific Publishers 2017
Online Access:https://www.scopus.com/inward/record.uri?eid=2-s2.0-85023780031&doi=10.1166%2fasl.2017.8237&partnerID=40&md5=9454e3fd52f829197aa05714e1cae368
Description
Summary:Ontology is an important element in the Semantic Web. It can represent scattered knowledge from unstructured text into a structured representation. However, extracting conceptual knowledge for the constructing of ontology is a time consuming task. Thus, the task is usually performed either automatically or in semi-automatic manner. The task of extracting knowledge or ontological elements is known as Ontology Learning. There are limited studies on Ontology Learning approach for Quran. Besides, most of the ontology constructions are manually done and focuses only on a few domains. Hence, there is a need to explore another domain and at the same time reducing the time to construct the ontology. Recent advances in Ontology Learning used NER method to extract the knowledge. The method is based on Information Extraction pipeline. This paper proposed a framework of Ontology Learning to construct ontology from Quran. The framework contains a method to extract ontological elements and eventually could be used to construct ontology from the extracted elements. This paper presents an initial experiment using NER method based on the framework. Results obtain from the experiment are in the form of concepts. Future work includes improvement of the NER method and construction of ontology for the selected domain. © 2017 American Scientific Publishers All rights reserved.
ISSN:19366612
DOI:10.1166/asl.2017.8237