Ron Snyder has over 30 years in the software industry with experience in both commercial and non-profit organizations. Ron is currently the Director of Research and Development for the JSTOR Labs team. He has been with JSTOR for 14 years holding various positions including software development manager and manager of website operations. Prior to coming to JSTOR Ron served 9 years in the military and worked in the aerospace industry for nearly 15 years in both engineering and management positions. Ron’s professional interests include information organization and discovery at scale. He was the principal architect of 2 popular JSTOR Labs projects, the Text Analyzer (https://www.jstor.org/analyze) and The Understanding Series (https://www.jstor.org/understand) which provide novel and innovative tools for content discovery and linking.
Why Wikibase? Why not?
In recent projects the JSTOR Labs (https://labs.jstor.org) team has jumped head first into the semantic web pool. Knowledge graphs and LOD are a great fit when the goal is to connect disparate data items which is a common theme in many of the Labs projects. We’ve explored a few solutions for our knowledge graph infrastructure but have settled on a self-hosted instance of Wikibase, the Mediawiki extension behind Wikidata.
In the first half of the presentation we will describe our rationale for the selection of Wikibase and share information on ongoing work to create a robust implementation of the full Wikibase stack using docker and kubernetes than can be scaled from a local laptop to a production grade cloud environment such as AWS. We will describe how we boot-strapped our Wikibase instance with a data model that both reflects the specific needs of our projects but also preserves a high-level of interoperability with Wikidata and other federated LOD repositories. At the conclusion of this part of the presentation the attendees will be able to access, install and run a fully functional version of a vanilla Wikibase configuration locally using the kubernetes setup described.
In the second part of the presentation we will explore a couple of specific use cases in which the Labs team uses our Wikibase setup to connect primary sources with entities in Wikidata and content in the JSTOR and Artstor repositories. We will provide a short demo of two projects in different stages of development, . Oone involving geospatial visualizations and another that uses a rich client app for interacting with audio-visual primary source content. We’ll share some initial lessons-learned in populating the knowledge base and using it as the primary data source for interactive user-facing web applications.