This masterclass on Graph Data Science offers a practical introduction to KGlab, an open-source Python library aimed at making graph data science accessible to everyone. Participants will explore different types of graphs and learn about the drawbacks of commercial solutions that often link computing and storage. Instead, the session promotes a decoupled approach. Attendees will see how to use KGlab to create knowledge graphs from diverse data sources, integrate ontologies and taxonomies, execute SPARQL queries, and apply various graph algorithms, including centrality and community detection. Additionally, the class covers probabilistic reasoning (PSL) and deep learning techniques for inference and data quality checks (Shackle). The key takeaway is the value of a flexible, integrated strategy that combines both sparse and dense representations, harnessing the strengths of various graph technologies for practical applications in sectors like manufacturing, finance, and pharmaceuticals.
Sign in to continue reading, translating and more.
Continue