Making Model Evolution Understandable to Domain Users

In complex domains, models such as knowledge graphs, undergo constant evolution to adapt to new data, insights, and requirements. Model evolution refers to the changes that occur over time between different versions of a model. While these updates are necessary to keep models accurate and relevant, they often introduce challenges for domain users who are not experts in the technical aspects of the changes. Ensuring that model evolution is comprehensible to these users is crucial for informed decision-making and continued trust in the systems.

 

This thesis focuses on how to make model evolution, specifically for knowledge graphs, understandable to domain users. The methodology explores graph evolution techniques or software update strategies, examining how changes can be formalised and represented in a human-readable format. A key aspect of the work will be defining what "understandable" means in this context and developing techniques that align with domain users' cognitive and interpretative capabilities.

Project Tasks:

  • Explore and analyse existing graph evolution and software update techniques.
  • Define the concept of "human-understandable" in the context of model evolution.
  • Apply formal evolution techniques to knowledge graphs and other models.
  • Develop methods for presenting model changes in a way that is accessible to non-technical users.
  • Evaluate the effectiveness of the proposed approach through user studies or case studies.
Published Oct. 23, 2024 9:37 AM - Last modified Oct. 23, 2024 9:37 AM

Supervisor(s)

Scope (credits)

60