1st Workshop on Beyond Attribute-Value Case Representation (BEAR)

Related to this year's conference theme, “CBR in a Data-Driven World”, this workshop addresses the challenges of coping with complex data in CBR. The BEAR workshop discusses research work, CBR applications, and systems where typical attribute-value case representations reach their limits and complex case representations such as object-oriented, textual, graph-structured, time-oriented (time series), hierarchical or hybrid representations need to be used. The application of such complex representations implies necessary (research) challenges. In general, these challenges affect all phases of the CBR cycle. For example, the complex case representation impacts the performance of similarity-based retrieval and adaptation. In addition, it affects the use of other AI methods integrated with CBR. In this workshop, participants shall present their research with complex case representations focusing on the general challenges and the impact on the CBR phases. The workshop aims to foster collaboration and exchange of ideas among researchers, developers, and others who use complex case representations that go beyond attribute-value case representations. 

Contributions should be in areas that include, but are not limited to, the following topics: 

• Representation, Authoring, Elicitation, and Visualization of Case Representations beyond Attribute-Value Representations (Text, Graphs, Workflows/Processes, Plans, Time Series Data, IoT Sensor Data, etc.) 

• Similarity-Based Retrieval Methods for Complex Cases 

• Learning of Similarity Measures 

• Adaptation Methods for Complex Case Representation (Transformational Adaptation as well as Generative Adaptation) 

• Methods for Explainability of Experience Reuse for Complex Case Representations 

• Evaluation of CBR Systems using Complex Case Representations 

• Methods for Supporting Users in the Revise-Phase 

• Tools and Techniques for the Retainment of Complex Cases 

• CBR Frameworks, Systems, or Applications supporting Complex Case Representations or Using Real-World Data