Association for Computing Machinery
The LOD cloud comprises of billions of facts covering hundreds of datasets. In accordance with the linked data principles, these datasets are connected by a variety of typed links, forming an interlinked \"Web of data\". The growing diversity of the web of data makes it more and more challenging for publishers to find relevant datasets that could be linked to, particularly in specialist domain-specific settings. This paper thus proposes a baseline method to automatically identify a list of public SPARQL endpoints whose content is deemed relevant to a local dataset based on queries generated from a local set of domain-specific keywords.