TY - GEN
T1 - Publishing E-RDF linked data for many agents by single third-party server
AU - Wang, Dongsheng
AU - Zhang, Yongyuan
AU - Wang, Zhengjun
AU - Chen, Tao
PY - 2017
Y1 - 2017
N2 - Linked data is one of the most successful practices in semantic web, which has led to the opening and interlinking of data. Though many agents (mostly academic organizations and government) have published a large amount of linked data, numerous agents such as private companies and industries either do not have the ability or do not want to make an additional effort to publish linked data. Thus, for agents who are willing to open part of their data but do not want to make an effort, the task can be undertaken by a professional third-party server (together with professional experts) that publishes linked data for these agents. Consequently, when a single third-party server is on behalf of multiple agents, it is also responsible to organize these multiple-source URIs (data) in a systematic way to make them referable, satisfying the 4-star data principles, as well as protect the confidential data of these agents. In this paper, we propose a framework to leverage these challenges and design a URI standard based on our proposed E-RDF, which extends and optimizes the existing 5-star linked data principles. Also, we introduce a customized data filtering mechanism to protect the confidential data. For validation, we implement a prototype system as a third-party server that publishes linked data for a number of agents. It demonstrates well-organized 5-star linked data plus E-RDF and shows the additional advantages of data integration and interlinking among agents.
AB - Linked data is one of the most successful practices in semantic web, which has led to the opening and interlinking of data. Though many agents (mostly academic organizations and government) have published a large amount of linked data, numerous agents such as private companies and industries either do not have the ability or do not want to make an additional effort to publish linked data. Thus, for agents who are willing to open part of their data but do not want to make an effort, the task can be undertaken by a professional third-party server (together with professional experts) that publishes linked data for these agents. Consequently, when a single third-party server is on behalf of multiple agents, it is also responsible to organize these multiple-source URIs (data) in a systematic way to make them referable, satisfying the 4-star data principles, as well as protect the confidential data of these agents. In this paper, we propose a framework to leverage these challenges and design a URI standard based on our proposed E-RDF, which extends and optimizes the existing 5-star linked data principles. Also, we introduce a customized data filtering mechanism to protect the confidential data. For validation, we implement a prototype system as a third-party server that publishes linked data for a number of agents. It demonstrates well-organized 5-star linked data plus E-RDF and shows the additional advantages of data integration and interlinking among agents.
KW - Data integration
KW - E-RDF
KW - Knowledge representation
KW - Linked data
KW - Semantic web
KW - Web service
U2 - 10.1007/978-3-319-70682-5_10
DO - 10.1007/978-3-319-70682-5_10
M3 - Paper in conference proceeding
AN - SCOPUS:85033791249
SN - 9783319706818
VL - 10675 LNCS
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 151
EP - 163
BT - Semantic Technology - 7th Joint International Conference, JIST 2017, Proceedings
PB - Springer
T2 - 7th Joint International Conference on Semantic Technology, JIST 2017
Y2 - 10 November 2017 through 12 November 2017
ER -