UPM Institutional Repository

Semantic triple ranking based on levenshtien reverse engineering approach


Citation

Yauri, Aliyu Rufai and Abdul Kadir, Rabiah and Azman, Azreen and Azmi Murad, Masrah Azrifah (2015) Semantic triple ranking based on levenshtien reverse engineering approach. Research Journal of Applied Sciences, Engineering and Technology, 11 (5). pp. 468-472. ISSN 2040-7459; ESSN: 2040-7467

Abstract

In sematic Web data are represented in Resource Description Framework (RDF) in triple format (Subject, relation, Object) and retrieved using structured query such as SPARQL. These structured queries require complex syntax to formulate. In view of this therefore, several approaches have been researched to enables semantic formulation of natural language to structure query. The process involves the representation of natural language query to structured triple format. However, dues complex nature of natural language, one natural language query may have more than one possible triple format; therefore an effective semantic triple ranking framework is needed for semantic triple ranking. In this study, semantic triple ranking mechanism is proposed. The approach is based on using levenshtien string matching algorithm a reverse engineering approach. The result of the proposed triple ranking has increased precision to 0.04 and recall 0.06.


Download File

[img] Text
Semantic triple ranking based on levenshtien reverse engineering approach.pdf
Download (84kB)

Additional Metadata

Item Type: Article
Divisions: Faculty of Computer Science and Information Technology
DOI Number: https://doi.org/10.19026/rjaset.11.1849
Publisher: Maxwell Science Publications
Keywords: Concept; Information retrieval; Predicate; Quran ontology; Semantic web; Triple
Depositing User: Ms. Ainur Aqidah Hamzah
Date Deposited: 17 Jun 2022 02:01
Last Modified: 17 Jun 2022 02:01
Altmetrics: http://www.altmetric.com/details.php?domain=psasir.upm.edu.my&doi=10.19026/rjaset.11.1849
URI: http://psasir.upm.edu.my/id/eprint/46260
Statistic Details: View Download Statistic

Actions (login required)

View Item View Item