Semantification of CEUR-WS with Wikidata as a target Knowledge Graph

For modern scholarly communication infrastructure, knowledge graphs are already ubiquitous. Some platforms of the publication lifecycle still struggle with catching up. In this paper, we use the example of the CEUR-WS publishing platform to show a viable approach to transition from a traditional HTML/PDF/text based environment to a single point of truth that separates the data and metadata storage from its presentation. This is possible using a public infrastructure such as Wikidata, which minimizes the maintenance effort and improves the publishing workflow. The CEUR Workshop Proceedings (CEUR-WS) publishing platform (https://ceur-ws.org/) has been introduced in 1995 as a means to publish proceedings of scientific workshops (and smaller conferences) in computer science. Technically HTML, PDF and a filesystem directory hierarchy are the core formats being used and delivery is via the HTTP and FTP protocols. There have been multiple attempts in the past to make the metadata of the CEUR-WS platform available for computer based analysis and querying. None of these attempts has been consistent and continuous so far. We report on the successful start of semantification of CEUR-WS with Wikidata as a target knowledge graph with the goal to achieve consistency and continuity for the future. The challenge was in handling the textual natural language description parts of the CEUR-WS content that is inherently still part of the semi-digitized approach of using HTML and PDF. We propose to have a better separation of concerns of metadata, display and storage and started implementing it.