CITS3005 Knowledge Representation

The UWA Book of Knowledge

This project is due on 5pm October 20, 2023 and is worth 40% of your final grade. You may choose to complete the project in a pair, or as an individual. If you complete the project as a pair, you must each submit an individual report. Each student should submit their work to cssubmit. If working in a pair, each student should submit the full set of files, but each file should clearly indicate both students who contributed to it.

Overview:

The UWA handbook contains critical high level information about all of the University's degree programs, majors and units. This project requires you to build a knowledge graph representing a fragment of that information, and providing an OWL ontology with some basic reasoning capabilities. The knowledge graph onlky need to contain information on units, and degrees. Courses, students, grades, staff, campus's etc are out of scope and can be represented by literals.

Deliverables:

You should deliver the following elements:

1. 25% A knowledge graph, implemented in python using rdflib. The data should be contained in a single file, with a python script provided to load the data into RDFLib and execute some SPARQL queries, including:
   • Find all units with more than 6 outcomes.
   • Find all level 3 units that do not have an exam, and where none of their prerequisites have an exam.
   • Find all units that appear in more than 3 majors.
   • Basic search functionality: Given a query string (eg "environmental policy"), can you find the units that contain this string in the description or outcomes?
2. 15% A SHACL Constraints File representing the following constraints:
   • Every prerequisite for a level X unit should have a level less than X
   • No unit should be its own prerequisite
   • No major should require more than 40 contact hours per week
   This should be submitted as a shapes graph and a python script for running validation.
3. 25% An OWL ontology describing the concepts in the graph, including: unit, major, outcome, prerequisite, assessment, contact. These should natural and useful queries over the knowledge graph, and should enforce, at least:
   • A prerequisite of a prerequisite is a prerequisite.
   • An outcome of a core unit is an outcome of a major
   • A required text of a core unit is a required text for a major
   This should be submitted as an owl file, with an python script for loading the ontology with the knowledge graph.
4. 25% A User Manual that:
   • Gives an overview of the schema, constraints and ontology rules.
   • Example queries, describing how to form the queries and interpret the output.
   • Instructions on how to add, update or remove data and rules to the knowledge graph.
   • Instructions on how to check consistency on the knowledge graph, and identify errors.
   The user manual maybe presented as a pdf file, or using HTML/Markdown.
5. 10% An individual report describing the process of building the project, including:
   • The design choices made in the project. Which options were considered and why di you make the choice you did?
   • What tools you used in the project, how effective were they, and what you would recommend people use in the future
   .
   • An estimate of time spent on the different tasks in the project.
   • If you worked in pairs, describe how work was divided, how effective you thought the collaboration was, and an honest assessment of the relative contribution of both team members.
   The individual report should be a pdf file.

Resources:

• RDFLib and Python for building the knowledge graph and executing SPARQL queries.
• pYSHACL for applying and validating constraints on the graph.
• owlready2 for representing the ontology and applying reasoning.

Logic and Logic Programming

Knowledge Representation and Reasoning
by Ronald Brachman and Hector Levesque
2004, Elsevier
Knowledge Representation and Reasoning is a standard text for knowledge representation in AI. It is useful, but not essential to have a copy. We will focus on chapters 1-5, and 9-12.

Simply Logical - Intelligent Reasoning by Example
by Peter Flach
1994, John Wiley
Simply Logical is a good basic introduction to reasoning with prolog. It is available as an e-book with executable examples.

Knowledge Graphs and Ontologies

Knowledge Graphs
by Aidan Hogan, Eva Blomqvist, Michael Cochez, Claudia d’Amato, Gerard de Melo, Claudio Gutierrez, Sabrina Kirrane, José Emilio Labra Gayo, Roberto Navigli, Sebastian Neumaier, Axel-Cyrille Ngonga Ngomo, Axel Polleres, Sabbir M. Rashid, Anisa Rula, Lukas Schmelzeisen, Juan Sequeda, Steffen Staab, Antoine Zimmermann
2021, Synthesis Lectures on Data, Semantics, and Knowledge
Knowledge graphs is a free online book giving a good overview of the theory and technologies of knowledge graphs.