OBDA Stream Access Combined with Safe First-Order Temporal Reasoning

Stream processing is a general information processing paradigm with different applications in AI. Most stream languages rely on the concept of a sliding window with a bag semantics, which is in order for relational streams but may lead to inconsistencies when applied on streams of assertions evaluated against a conceptual model. Our approach uses a different semantics based on ABox sequencing. The query language provides an expressive first order temporal logic for inter-ABox reasoning. Safety conditions tame the expressiveness so that a meaning preserving transformation of the query to backend queries on the sources as foreseen in the OBDA paradigm is guaranteed.