A New Method for Coreference Resolution in the Web of Linked Data Based On Machine Learning
Main Article Content
Abstract
Web of Linked Data forms a single, globally distributed data space. Establishing RDF links between overlapping but separately constituted RDF datasets still represents one of the most important challenges to achieve the vision of the Web of Linked Data. In Linked Data environment, an object is likely to be denoted with multiple URIs by different data providers. Object coreference resolution is to identify “equivalent†URIs that denotes the same object. One of the most important types of RDF links are “Identity Linksâ€, which point at coreferent objects. By common agreement, Web of Linked Data uses owl:SameAs predicate to state identity links. Driven by the Linking Open Data (LOD) project, millions of URIs have been explicitly linked via owl:sameAs, but potentially coreferent ones are still considerable. Coreference resolution and data linking often relies on fuzzy similarity functions comparing relevant characteristics of objects in the considered datasets and manually tuned metrics for estimating similarity between objects. In this paper, we describe an approach for object coreference resolution in Linked Data, which relies on supervised learning and support vector machines. We propose to employ different similarity functions and combine them with a learning scheme. Initial experiments applying this approach to public datasets have produced encouraging results.
Â
Keywords: Coreference Resolution, Data Interlinking, Linked Data, Semantic Web, SVM.
Downloads
Article Details
COPYRIGHT
Submission of a manuscript implies: that the work described has not been published before, that it is not under consideration for publication elsewhere; that if and when the manuscript is accepted for publication, the authors agree to automatic transfer of the copyright to the publisher.
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgment of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgment of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work
- The journal allows the author(s) to retain publishing rights without restrictions.
- The journal allows the author(s) to hold the copyright without restrictions.