The majority of available datasets in open government dataare statistical. They are widely published by various governments to beused by the public and data consumers. However, most open data por-tals do not provide the five-star Linked Data standard datasets. Thepublished datasets are isolated from one another while conceptually con-nected. Through this paper, a knowledge graph is constructed for thedisease-related datasets of a Canadian government data portal, NovaScotia Open Data. We leverage the Semantic Web technologies to trans-form the disease-related datasets into the Resource Description Frame-work (RDF) standard and enrich them with semantic rules. An RDFdata model using the RDF Cube vocabulary is designed in this work todevelop the graph that adheres to best practices and standards, allowingfor expansion, modification and flexible re-use 3. The study also discussesthe lessons learned during the cross-dimensional knowledge graph con-struction and integrating open statistical datasets from multiple sources.