Title: Ontology-based metadata matching for emergency decision-making using MapReduce

Authors: Li Zhu; Wei Hu

Addresses: School of Economics and Management, Nanjing University of Information Science and Technology, Nanjing 210044, Jiangsu, China ' State Key Laboratory for Novel Software Technology, Department of Computer Science and Technology, Nanjing University, Nanjing 210023, Jiangsu, China

Abstract: Ontology matching techniques can help improve the accuracy of emergency decision-making on heterogeneous data. In this paper, we propose a practical approach to leverage ontology and MapReduce for matching metadata of emergency supplies. We use this approach in the domain of emergency food. Specifically, by extending the AGROVOC ontology developed by the Food and Agriculture Organisation (FAO) of United Nations, ontological descriptions of food metadata are established. Then, based on the classification of emergency functionality, the instances of food metadata within the same category are matched in a two-stage TF-IDF fashion, which is further improved with the MapReduce framework for efficient parallel computation. Our experiments on grain, meat and milk metadata retrieved from several e-commerce websites demonstrate the good performance of the proposed approach. Additionally, case study is provided to illustrate the potential use of our approach.

Keywords: ontology matching; MapReduce; TF-IDF; emergency decision making; food metadata; emergency supplies; emergency food supplies; grain metadata; milk metadata; meat metadata; emergency management; natural disasters; disaster management.

DOI: 10.1504/IJMSO.2015.070821

International Journal of Metadata, Semantics and Ontologies, 2015 Vol.10 No.2, pp.75 - 83

Received: 08 Feb 2014
Accepted: 19 Oct 2014
Published online: 28 Jul 2015 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article