Providing packages of relevant ATM information: An ontology-based approach

ATM information providers publish reports and notifications of different types using standardized information exchange models. For a typical information user, e.g., an aircraft pilot, only a fraction of the published information is relevant for a particular task. Filtering out irrelevant information from different information sources is in itself a challenging task, yet it is only a first step in providing relevant information, the challenges concerning maintenance, auditability, availability, integration, comprehensibility, and traceability. This paper presents the Semantic Container approach, which employs ontology-based faceted information filtering and allows for the packaging of filtered information and associated metadata in semantic containers, thus facilitating reuse of filtered information at different levels. The paper formally defines an abstract model of ontology-based information filtering and the structure of semantic containers, their composition, versioning, discovery, and replicated physical allocation. The paper further discusses different usage scenarios, the role of semantic containers in SWIM, an architecture for a semantic container management system, as well as a proof-of-concept prototype. Finally the paper discusses a blockchain-based notary service to realize tamper-proof version histories for semantic containers. •Semantic containers build on ontology-based filtering of ATM information.•Uniform represention of information needs and information packages as containers.•Matching of information needs and available information packages is automated.•Container versioning facilitates reconstruction of past information states.•Replicated container allocation facilitates resynchronization after network failures.