Improvement of Supervised Machine Learning Methods in the Web of Linked Data by using Available Owl:sameAs Links in the LOD Cloud

Main Article Content

Leila Namnik
Mehran Mohsenzadeh, Mashalla Abbasi Dezfouli


The web of Linked Data is characterized by linking structured data from different sources using equivalence statements, such as Owl:sameAs, as well as other types of properties. Object coreference resolution is to identify “equivalent†URIs that denotes the same object. Owl:sameAs links will be established between coreferent URIs, identified by coreference resolution methods. In our previous work, we described an approach for object coreference resolution in the Linked Data environment which relied on standard supervised machine learning methods and support vector machines (SVMs). We proposed to employ different similarity functions and combined them with a learning scheme. In this paper, we extend our previous approach by using existing Owl:sameAs links already exist in the web of Linked Data by Linking Open Dataset (LOD) project. By using these links, we could substitute the process of manually labeling training examples in the learning model with an automatic one. We evaluate our approach on common datasets and obtain encourage results, that offer performance comparable with state-of-the-art non learning based systems on these datasets.


Keywords: Coreference Resolution, Data Interlinking, Owl:sameAs, Linked Data, Semantic Web, SVM.


Download data is not yet available.

Article Details