The SILKNOW knowledge graph

SILKNOW is a research project that aims at improving the understanding, conservation and dissemination of the European silk heritage from the 15th to the 19th century. This paper presents the SILKNOW knowledge graph (KG) that lies at the center of the application of Semantic Web technologies and computing research to the needs of museums and every other user of this knowledge. The underlying data model is based on CIDOC-CRM and data mappings which are realised and implemented with conversion tools developed for SILKNOW. The full integration pipeline consists also of our own crawling software to retrieve the original data from both public sources and project partners. We developed an API access for the KG and created the exploratory search engine ADASilk on top of it. Finally, we present how we apply automatic image and text analysis to predict missing metadata in the knowledge graph.