eCite Digital Repository

Mining semantic association rules from RDF data


Barati, M and Bai, Q and Liu, Q, Mining semantic association rules from RDF data, Knowledge-Based Systems, 133 pp. 183-196. ISSN 0950-7051 (2017) [Refereed Article]

Copyright Statement

Copyright 2017 Elsevier B.V.

DOI: doi:10.1016/j.knosys.2017.07.009


The Semantic Web opens up new opportunities for the data mining research. Semantic Web data is usually represented in the RDF triple format (subject, predicate, object). Large RDF-style Knowledge Bases contain hundreds of millions of RDF triples that represent knowledge in a machine-understandable format. Association rule mining is one of the most effective techniques for detecting frequent patterns. In the context of Semantic Web data mining, most existing methods rely on users intervention that is time-consuming and error-prone due to a large amount of data. Meanwhile, rule quality factors (e.g. support and confidence) usually consider knowledge at the instance-level. Namely, these factors disregard the knowledge embedded at the schema-level. In this paper, we demonstrate that ignoring knowledge encoded at the schema-level negatively impacts the interpretation of discovered rules. We introduce an approach called SWARM (Semantic Web Association Rule Mining) that automatically mines Semantic Association Rules from RDF data. The main achievement of SWARM is to reveal common behavioural patterns associated with knowledge at the instance-level and schema-level. We discuss how to utilize knowledge encoded at the schema-level to add more semantics to the rules. We compare the semantic of rules discovered by SWRAM with one of the latest approaches in this field to show the importance of considering schema-level knowledge. Initial experiments performed on RDF-style Knowledge Bases demonstrate the effectiveness of the proposed approach.

Item Details

Item Type:Refereed Article
Keywords:Semantic Web data, association rule mining, ontology, knowledge discovery
Research Division:Information and Computing Sciences
Research Group:Computer vision and multimedia computation
Research Field:Pattern recognition
Objective Division:Information and Communication Services
Objective Group:Information systems, technologies and services
Objective Field:Application software packages
UTAS Author:Bai, Q (Dr Quan Bai)
ID Code:140662
Year Published:2017
Web of Science® Times Cited:13
Deposited By:Information and Communication Technology
Deposited On:2020-09-01
Last Modified:2020-10-23

Repository Staff Only: item control page