11 July 2021
Data Model vs Ontology Development – a FIBO perspective
by Mike Bennett, Hypercube & EDM Council
- 1. The FIBO Journey: To Ontology and Beyond! Connected Data London Savoy Place, London 16 November 2017 Mike Bennett 1
- 2. Outline • FIBO Motivations • Early Explorations • Developing a Concept Ontology • The OWL Experience • Refining he principles for Concept Modeling 2
- 3. FIBO Motivations 3
- 4. FIBO Motivations 4 ? ? ? ? ?
- 5. FIBO Motivations 5 Common ontology Shared business meanings
- 6. FIBO Motivations 6 Common ontology Shared business meanings Validated by business Expressed logically
- 7. Early Explorations 7
- 8. 8 Possible classes of Thing
- 9. 9 Example “Thing”: Equity • Real world definition of Equity: "An equity is a financial instrument setting out a number of terms which define rights and benefits to the holder in relation to their holding a portion of the equity within the issuing company".
- 10. 10 What is an Equity? Or to put it another way… Equity Equity security Instrument Terms Financial Instrument Is a kind of Has rights defined in In relation to
- 11. 11 What is an Equity? Using OWL to define the classes of real things in the world, and the facts about those things Modeled in TopBraid Composer
- 12. 12 Financial Semantics in OWL • Pizza approach • “Everything is a Thing” • What about common terms? • accounting terms for equity, debt, cashflow • Places, time concepts • Legal terms (securities are contracts) • Better partitioning needed
- 13. The Semantic Web • Web Ontology Language • Based on Subject-Verb-Object “Triples” • Widely used • Protégé tool • Experiment: Ingest a logical data model into OWL • Result: a logical data model in OWL • Syntax is not semantics! 13
- 14. Developing a Concept Ontology 14
- 15. The FIBO Moment • Previous standardization efforts at message and data levels • Arguments over terms • Atkin: “What if we considered the concepts without worrying about the words people use?” • Sudden outbreak of peace! 15
- 16. Financial Industry Business Ontology Semantics Repository Industry Standards XLS Boxes & LinesUser Commitments Original Content ISO 20022 FpML XBRL SemWeb OWL constructs ODM EA UML Tool MDDL enhancements for readability Theory of meaning creates SME Reviews Tweaks for Tool support RDF/OWL 16
- 17. FIBO: Scope and Content Upper Ontology FIBO Foundations: High level abstractions FIBO Contract Ontologies FIBO Pricing and Analytics (time-sensitive concepts) Pricing, Yields, Analytics per instrument class Future FIBO: Portfolios, Positions etc. Concepts relating to individual institutions, reporting requirements etc. FIBO Process Corporate Actions, Securities Issuance and Securitization Derivatives Loans, Mortgage Loans Funds Rights and Warrants FIBO Indices and Indicators Securities (Common, Equities) Securities (Debt) FIBO Business Entities FIBO Financial Business and Commerce 17
- 18. The OWL Experience 18
- 19. Two Ontological Traditions: 19 Semantic WebApplied Ontology FIBO The science of meaning Meaning expressed in formal logic Presented in the “Language of the business” Formally grounded in legal, accounting etc. abstractions Use case specific ontologies Richer internal logic Focus on data Optimized for operational functions (reasoning; queries) Addition of rules Mapping to other OWL ontologies
- 20. FIBO Development & Feedback Ecosystem FIBO CORE: RDF / OWL is the system of record for FIBO (everything needed for inference processing) FIBO Vocabulary: The FIBO business conceptual model expressed in SKOS (everything needed for the unification of data across repositories) FIBO OMG: Standards partner with EDMC for visualizing FIBO in UML (everything needed for expressing FIBO as diagrams) FIBO.Schema.org FIBO aligned to the Schema.org community financial data used for mapping existing web pages to FIBO FIBO CORE RDF / OWL FIBO Vocabulary SKOS – RDF/S FIBO OMG FIBO.Schema.org UML / SIMF Generated Generated Aligned Industry Feedback 20
- 21. Semantic Web Applications Swap1001 Leg 1 Leg 2 10000000 notional notional LIBOR 3.5% Fixed Float IR Swap LEI5001 LEI7777 Trader LLCAcme Inc identifies identifies USD currency Interest Rate Swap 21 10000000 USD currency Swap FloatingRateLeg Inferred Leg1 is inferred to be a FloatingRateLeg because any leg tied to an index is semantically defined as floating Inferred FixedRateLeg Inferred Leg2 is inferred to be a FixedRateLeg because any leg tied to an interest rate is semantically defined as fixed LEI LEI Business EntityBusiness Entity Swap is inferred to be a Fixed-Float IR Swap because one leg was inferred to be fixed and one leg was inferred to be floating fulfilling the definitions in the ontology Inferred Data for an undefined Swap Contract before semantic reasoning performs classification and identification type type type type An interest rate swap in which fixed interest payments on the notional are exchanged for floating interest payments. Human Facing Definition Swap_Contract and hasLeg FixedRateLeg and hasLeg FloatingRateLeg Machine Facing Definition Fixed Float IR Swap (Ontology) Semantic reasoning Semantic reasoning Semantic reasoning1 2 3 isTradingWith isTradingWith is a new property relationship that is inferred based on a semantic rule and can be queried Semantic reasoning4 fixedRateindex • Semantic Operational Processing Reasons over Data to Infer Classifications and Relationships David Newman, Wells Fargo
- 22. Properties with No Domain or Range 22
- 23. Two Approaches to Meaning 23 Rosetta Stone Mayan Language • Existence of already-understood terms enabled translation • Semantics grounded in existing sources • No existing common language to enable translation • Translation was possible only from internal consistency of concepts
- 24. Internal Correspondence Semantics • Graph has logical relations between elements • These correspond to the relations between things in reality • Automated reasoning checks the “deductive closure” of the graph for consistency and completeness 24 Mayan Language
- 25. 25 • Directed Graph • The meaning at each node is a product of its connections to other nodes • Semantically grounded at certain points in the graph Semantic Networks Foundational Semantics Rosetta Stone
- 26. Which is Which? • Foundational semantics: • External grounding of concepts based on things outside the ontology • Typically social constructs, commitments, legal and other primitives • Each is the “simplest kind of thing” of that type • Internal Correspondence Semantics • The deductive closure of the whole graph is where meaning comes in • Logic in the graph corresponds to relationships among things in the world • Use Foundational Semantics for a business concept ontology 26
- 27. Using OWL Business Conceptual Ontology (CIM) Operational Ontology (PSM) Extract and Optimise The Language Interface Business Technology 27
- 28. Using OWL: Datatypes Business Conceptual Ontology (CIM) Operational Ontology (PSM) Extract and Optimise The Language Interface Business Technology Data types Data types Platform specific matter • So what’s that about? • An OWL based conceptual ontology plus data seems to be a physical design artifact • But it is still conceptual in that it represents business concepts • Provided those concepts are expressed in data • There are real things, the definitions of which are not based on data!28
- 29. Refining the principles for Concept Modeling 29
- 30. The Three Elements Ontology Lexicon Data 30
- 31. Dimensions of a Model 31 Formalism Application Model Theoretic Relation (grounding) MODEL e.g. First Order Logic e.g. Business domain (business process etc.) e.g. Messaging Level
- 32. 32 Development Lifecycle for Data Level (from Zachman) Data Function 0 Scope (contextual) Things relevant to the business Set of business processes 1 Business Model (conceptual) Semantic Model Business Process Model 2 System Model (logical) Logical Data Model Logical Design 3 Technology Model (physical) Physical Data Model Physical Design 4 Detailed Representation Data definition Program
- 33. 33 Development Lifecycle for Data Level (from Zachman) Data Function 0 Scope (contextual) Things relevant to the business Set of business processes 1 Business Model (conceptual) Semantic Model Business Process Model 2 System Model (logical) Logical Data Model Logical Design 3 Technology Model (physical) Physical Data Model Physical Design 4 Detailed Representation Data definition Program
- 34. 34 This is not a more abstract model of the solution… Conceptual Ontology Logical Data Model (PIM) Physical Data Model (PSM) Realise Implement The Language Interface Business Technology
- 35. 35 This is not a more abstract model of the solution… Conceptual Ontology Logical Data Model (PIM) Physical Data Model (PSM) Realise Implement The Language Interface Business Technology It’s a concrete model of the problem!
- 36. The Semiotic Triangle (Peirce) Concepts Signs Real World Objects 36
- 37. Concepts Diagram: Jim Odell 37
- 38. Semiotic Rhombus Extensions Signs Real World Objects Intensions Concepts 38 • Separate intension and extension • Extension can happen one, many or no times • The ontology is the intensional model of meaning • Matters of ontological commitment to things in the world are based on usage of the ontology
- 39. Things Information Type A set specification for a kind of Independent Thing that generalizes all towers (e.g., “a tall narrow structure”) A set specification for a kind of Dependent Continuant that is a record structure containing tower observations (e.g., a “TOWER” table or a “#Tower” class) Sets One of many sets of independent things that generalize all towers One of many sets of dependent continuant record structures containing tower observations (e.g., in that database there) Member A member of zero or more sets of all towers (E.g., the actual one we call the “Eiffel Tower”) A member of one or more sets of record structures containing tower observations (E.g., one that represents the actual Eiffel Tower) “#tower123”Represents Introducing the Data Dimension Jim Logan, NoMagic 39
- 40. Data Delta 40 Things the data is about Data-focused Extensions about represents ẟD
- 41. Data Delta: ẟ => 0 41 Things the data is about about ẟD => 0
- 42. Examples: Logical (data-friendly) Intensions • Meaning of Bank: framed in terms of legal capabilities and rights • Logical intension: presence of banking license? • Ownership and Control • Confer certain rights and involve certain capabilities • These are social constructs not data • In general: Data surrogate for real thing • Look for signatures in data that imply the presence of real world, identifying matter • Frame the necessary conditions for membership of a class (in a logical ontology) in terms of what would be found (true) in data when the class of thing is there • Inference as distinct from meaning in the original sense • From the data you can infer that a thing exists in reality • Real meaning – by definition mostly does not rely on data! 42
- 43. Summary • FIBO is available at https://spec.edmcouncil.org/fibo • Production: Optimied for Semantic Web / Reasoning • Development: Extensive industry content, optimized for common semantics • Users should apply model theoretic thinking in making use of these ontologies • Inference processing: use optimized ontologies • Mapping, integration, NLP: use foundational semantics 43
