cmput696 is a project-based course where students will be exposed to methods for the creation and curation of knowledge graphs and some of their applications. There are no formal prerequisites, but students are expected to have basic familiarity with natural language processing, query languages, machine learning and embeddings.

Learning Objectives

This course pursues the following objectives: exposing students to state-of-the-art methods in knowledge graphs; contrasting SPARQL-style query answering and text processing methods for answering questions; offering hands-on programming opportunities on real open datasets; improving the communication skills of students.

Topics:

1. Knowledge Graphs
   • NELL
   • YAGO, DBpedia and Wikidata
   • Google's KG and other commercial KGs
2. KG construction:
   • Manual: Wikidata
   • Automatic: NELL
   • Semi-automatic: YAGO and DBpedia
3. Information extraction and KGs
   • Entity detection and disambiguation
   • Relation extraction
   • Universal Schemas
4. KG embeddings
   • Geometric embeddings
   • Semantic embeddings
   • Applications: link prediction, clustering, classification

There is no textbook for the course. The reading list will be updated throughout the term.