Global Trustee and Fiduciary Services News and Views Issue 50

Prime, Futures and Securities Services | Financial Innovation 64 SBVR is being used by major banks to help map regulatory concepts onto business concepts. Ontologies are being used as metadata models hosted in RDF triple stores as knowledge bases for data extracted from heterogonous relational data stores and other sources. This semantic approach to data virtualisation uses SPARQL (W3C semantic query language) to field federated queries over the distributed metadata/data in relational data stores. The operational data stays where it is, with the data of interest returned from multiple data sources, integrated using the ontologies (as metadata models), with further analysis and processing performed in an RDF triple store. This approach takes on an AI dimension when inferencing engines or reasoners are employed to add knowledge to a knowledge base. A semantic reasoner or rules engine consists of algorithms that infer logical conclusions from a set of asserted axioms or facts expressed in RDF/OWL. From a data perspective, previously unknown or unrecognised relationships across heterogeneous data sets can be asserted, thus adding more knowledge. Successful ontology- based solutions already exist in a wide variety of domains from defence and intelligence to capital markets to regulatory compliance. From a business perspective, this approach enables regulatory semantics (vocabularies and rules underpinning regulatory provisions and compliance imperatives) to be mapped to business semantics (vocabularies and rules that express business policies, operational standards, risks and controls through to metadata repositories/data dictionaries). Figure 3 presents a stylised model of the regulatory compliance value chain. It identifies the key areas or activities that are ripe for digitisation through regtech. However, we argue that to address the Tower of Babel problem, an enterprise semantic bus or layer is required that acts as a repository of a firm’s “common language”. At a human level, this should be expressed as sets of linked vocabularies and rules captured using the Object Management Group’s SBVR. What’s required are regtech applications that persists SBVR-like vocabularies and rules in RDF (and XML), thereby creating an Figure 3 Regulatory compliance value chain Regulator Drafting and Publishing Regulations Regulatory Vocabulary and Rules Law Professional Services and GRC Interpret and Disambiguate Compliance and Risk Vocabulary and Rules Analyse and Document Business Vocabulary and Rules Advise on Obligations etc. Mega-Data and Algorithms Data Governance and IT Draft Policies Draft Standards and Procedures Manage Risks and Controls Compliance and Risk Second line of defence Business First line of defence Digital Regulation Digital Regulatory Compliance