The field of knowledge graphs is moving towards increased integration and interoperability, with a focus on enabling polyglot access to diverse data sources. Recent developments have highlighted the benefits of combining different data management approaches, such as Linked Data and labelled property graphs, to support more comprehensive and flexible knowledge representation. The use of large language models (LLMs) is also becoming more prevalent, particularly in tasks such as dataset information extraction and knowledge graph question answering. Additionally, there is a growing emphasis on addressing the quality issues and pitfalls that exist in current knowledge graph datasets, with a view to creating more reliable and challenging benchmarks for evaluating knowledge graph-based systems. Noteworthy papers include:

• A paper introducing a framework for mapping RDF data to semantically equivalent LPG formats, enabling polyglot access to knowledge graphs.
• A paper presenting HuggingKG, a large-scale knowledge graph built from the Hugging Face community for ML resource management, and HuggingBench, a multi-task benchmark for IR tasks.
• A paper proposing ChatPD, an LLM-driven paper-dataset networking system that automates dataset information extraction and constructs a structured paper-dataset network.
• A paper introducing KGQAGen, an LLM-in-the-loop framework for systematically resolving pitfalls in KGQA datasets and producing challenging and verifiable QA instances.