Tag
The paper proposes Semantic Bayesian World Models (SBWMs) to bridge the gap between knowledge graphs' crisp assertions and foundation models' probabilistic reasoning, enabling a shared, evolving web of beliefs for unified reasoning under uncertainty.
The paper introduces AVA, a framework to evaluate NLP embeddings' ability to capture ontological reasoning, finding significant limitations and challenging the assumption that strong NLP performance translates to Semantic Web competence.
Introduces NL2SHACL-Bench, a benchmark suite for translating natural language requirements into SHACL shapes, evaluating four state-of-the-art LLMs and showing they struggle with semantic equivalence for complex patterns.
This paper presents FDD-ON, a modular ontology for representing VAV HVAC system components, faults, symptoms, and impacts to enable interoperable fault detection and diagnostics applications.
The article argues that as AI agents proliferate, a new layer of the internet based on knowledge graphs will emerge, reviving Tim Berners-Lee's Semantic Web concept to provide structured highways for agent interaction and reasoning.
This paper introduces a neuro-symbolic meta-policy for temporal knowledge-graph memory in partially observable reinforcement learning, combining RDF-based graph representations with symbolic memory management heuristics to achieve inspectable and adaptive control.
A guide explaining how to add JSON-LD structured data to personal websites for better SEO and richer link previews, with fundamentals and copy-paste examples for common schema types.
BLINKG is a benchmark designed to evaluate the mapping capabilities of Large Language Models (LLMs) in constructing Knowledge Graphs from heterogeneous data sources. It provides a standardized framework to assess how effectively LLMs establish correspondences between data schemas and ontology concepts.
This extended paper revisits Semantic Web Services insights for Knowledge Graphs, proposing a four-dimensional formal framework and an Agentic Affordance Profile (AAP) to enable principled KG selection, composition, and failure diagnosis at agent planning time.
arXiv reports on its ongoing HTML Papers project, highlighting improved conversion fidelity, corpus-scale HTML conversion reaching 75% error-free rate, initial MathML 4 Intent annotations for accessible speech, and a Rust port of LaTeXML to reduce costs.
This thesis proposes a modular framework for scalable uncertainty reasoning in knowledge graphs, addressing imprecise attribute values, probabilistic triple existence, and incomplete schema through tailored algebraic, logical, and geometric techniques.