• Reading:
  • AIMA: Chapter 2, Section 2.4.7, and Chapter 10, Sections 10.1-10.2. Focus on atomic, factored, and structured representations; ontological engineering; categories and objects; category hierarchies; inheritance; and the trade-off between expressive representations and reasoning complexity.
• Technical preparation:
  • Open today's notebook and run all setup cells and the first code cell in Segment 1 to verify that the notebook works in your environment. The output should display information about two artworks in three different formats.

• 08:30 – MS Forms
• 10:00 – MS Forms

1. A museum stores each artwork as one text description. What questions become difficult to answer?
2. Two artworks were created by the same person. Should the creator's name be copied into both records?
3. If every painting is a cultural object, must we record both categories for every painting?
4. Should a museum knowledge model contain every known fact about an artwork?

Notebook: HCAI_Lab_Knowledge_Representation.ipynb (Work in Google Colab or download the .ipynb file to local environment)

Work in groups of 2-4. Every group member must be able to explain the results.

Work in groups of 2-4 to curate a small digital exhibition using cultural heritage data from Wikidata.

Choose your own theme and audience, then adapt the SPARQL templates from the STARTER NOTEBOOK to find suitable objects. Your query should use at least two meaningful relationships, such as:

artwork -> creator -> place of birth
artwork -> collection -> museum location
artwork -> depicts -> person or place

Select 5-10 exhibits and justify their inclusion. Do not accept every query result automatically. Identify at least one rejected, ambiguous, surprising, or potentially missing object.

Next, change one modelling decision and compare the results. How did the selected graph path affect the exhibition? Which version better matches your theme and audience? What does each representation reveal or hide?

Start the challenge during the lab and continue it as part of the self-study regarding classes. Selected groups may share their exhibitions and modelling choices at the next lab, and effective solutions may receive bonus points.

• Ontology Development 101 – A practical guide to defining scope, classes, properties, hierarchies, and competency questions
• Protégé – A free ontology editor for modelling classes, relationships, individuals, and logical constraints.
• OWL 2 Primer – An accessible introduction to more expressive ontology modelling and automated inference.

• ConceptNet – A large semantic network representing everyday knowledge used in commonsense reasoning.
• Commonsense knowledge in ConceptNet and word associations – A research perspective on how different graph-based resources represent implicit everyday knowledge.

• OBO Foundry – A collection of interoperable ontologies used across biology, medicine, and scientific data integration.
• SAREF – Ontologies for connecting smart devices, sensors, measurements, buildings, energy systems, and IoT services.
• Google Enterprise Knowledge Graph – See how organizations reconcile entities and connect information stored across heterogeneous data sources.