Developing NLP Solutions for Automatic Knowledge Graph Construction, Part 1 | Panos Alexopoulos

This masterclass delves into the automation of knowledge graph construction through Natural Language Processing (NLP). In the first half, participants will explore the fundamentals of knowledge graphs, including their key features and the challenges they face in terms of semantic accuracy, explicitness, and broader acceptance. The latter half focuses on entity mining with spaCy, an open-source NLP toolkit. Participants will learn to utilize pre-trained models, develop custom extraction rules, and refine models with additional examples to enhance accuracy and broaden entity recognition. For instance, the instructor will showcase how to improve the identification of "product" entities by introducing tailored rules and training data. Ultimately, this course aims to empower attendees with practical techniques for constructing and scaling knowledge graphs from raw text data.