392192 Actionable Knowledge Representation (V) (WiSe 2024/2025)

This course deals with the idea of bringing knowledge into applications to support users in daily life. It therefore covers topics on how knowledge can be represented to be machine-understandable, how knowledge can be acquired from different sources (including Web scraping) and how such different knowledge chunks can be linked. It will further discuss how to reason about knowledge and how different agents like websites, AR applications or robots can use knowledge to support users in their daily life.
All exercises will be available in platform-independent jupyter notebooks based on python and have low software requirements.

Throughout this course the students work on a knowledge representation project of their choice alone or in small groups (1-3 students). They focus on defining the fitting key terms & concepts, creating and querying an ontology and creating a web-based application for accessing and using the resulting knowledge representation. The different parts of the project are graded and make up the course result.

Learning Goals: The students ...
... get to know different types of knowledge
... learn why knowledge representation is important for agent applications
... are able to represent knowledge for agent applications
... get to know different techniques to acquire knowledge from the web
... can apply knowledge acquisition to extract relevant web knowledge
... learn how to extend acquired knowledge for agent applications
... are able to query their acquired knowledge
... know how different agents can query the knowledge base
... can generate a knowledge base that can be queried and used in an agent application