What happened
This is a thesis describing methods to automatically extract structured knowledge (called Knowledge Graphs) from massive collections of messy, unorganized text — like news articles, medical records, and research papers. The work combines natural language processing and AI techniques to turn scattered information into organized databases that computers can reason about and people can query, tested on three real domains: tracking digital transformation discussions, mapping construction research trends, and extracting medical cause-and-effect relationships from health records.
Why it matters
As organizations drown in unstructured text data they cannot easily search or analyze, automated methods that can reliably extract and organize knowledge at scale could unlock value trapped in millions of documents — but this paper is a single thesis contribution with no evidence of deployment or measurable real-world impact yet.