Building knowledge graphs with maplib

In data science, analysts have been using iterative environments like Jupyter Notebook and RStudio to perform exploratory data analysis. Such support has been missing from the world of Knowledge Graphs… until now!

maplib allows incremental, interactive knowledge graph construction using dataframes in Python and Jupyter Notebooks, and features immediately-available SPARQL querying of the knowledge graph being constructed, so that the graph can be validated and enriched.

We will survey the knowledge graph construction landscape, both in terms of tools and languages, and explain why it was necessary to build something else. We will then do a short workshop where we look at a couple of use cases of maplib:

• Exploratory Graph Data Science
• Incrementally building a robust graph mapping for an industrial use case

My goal is to quickly get you up and running, exploring the power of knowledge graphs using maplib!

maplib is one of the results of my PhD, and I am working to commercialize it in a startup called Data Treehouse ( https://www.data-treehouse.com/ ), but the core features of maplib are, and will remain open source with an Apache 2.0 license ( https://github.com/DataTreehouse/maplib ).

Speaker: Magnus Bakken
Teaser: https://ieeexplore.ieee.org/document/10106242