Project Scope Semantic interoperability Community of Practice Enablement

Project “Scope” Semantic interoperability Community of Practice Enablement for: – Federal Health Architecture (FHA) – National Health Information Network (NHIN) DRM Semantic Technologies Profile Pilot by – Ontolog Forum, – Stanford Medical Informatics, – SICo. P 1

Outline • • Participants Purpose Topics Demonstrations – Protégé based EON – Time Representations • Next Steps 2

Participants • ONTOLOG Forum: – an open, international, citizen-centric, virtual community of practice working on business domain ontologies. – http: //ontolog. cim 3. net • Stanford Medical Informatics (SMI): – an interdisciplinary academic and research group within the Department of Medicine in the Stanford University School of Medicine – brings together scientists who create and validate models of how knowledge and data are used within biomedicine. – http: //www. smi. stanford. edu/ • Semantic Interoperability Community of Practice (SICo. P): – chartered by the CIO Council’s Best Practices Committee for the purpose of achieving "semantic interoperability" and "semantic data integration" – focused on the government sector and to make the Semantic Web operational. – http: //web-services. gov 3

Purpose • to see how ontologies can help architects define their architecture • to understand how ontologies will help architects drive interoperability in ways that: – avoid remedial work down the road – avoid “false starts” 4

Specific Objectives • Provide Semantic Technology Profiles (STP) for the Data Reference Model (DRM). – See http: //colab. cim 3. net/cgibin/wiki. pl? Data. Reference. Model • Support the Federal Health Architecture’s (FHA) Architecture Development Methodology (ADM) and the Architecture Peer Review Group (APRG). – See http: //fha. mitre. org/ecommunities/login. jsp (password required) • Support the National Health Information (NHIN) Network Request for Information (RFI) Review. – See http: //www. hhs. gov/healthit/ 5

Main Topics (by Slide numbers) • Ontologies for Semantic Interoperability in Enterprise Architecture (Slide # 8) • Formal Taxonomies as Ontologies (Slide # 9) • Ontology-Driven Information Systems (Slide #10) • Categories of Ontologies (Slides # 11 -12) • FEA Reference Model Ontology (FEA-RMO) (Slide #13) • Business Reference Model (BRM) Taxonomy (Slide #14) • Health Domains Taxonomy (Slides #15 -16) • The Role of Ontologies in the Electronic Health Record (Slide #17) 6

Ontologies for Semantic Interoperability in Enterprise Architecture GIGO/minis/micros www / Netscape Web services OWL Age of Programs Age of Proprietary Data Age of Open Data Program-Data Text, Office Docs Databases (proprietary schema) HTML, XML (open schema) Namespaces, Taxonomies, RDF 1994 - 2000 - 2003 1945 -1970 Procedural Programming “Data is less important than code” 1970 - 1994 Age of Open Metadata Age of Semantic Models Ontologies & Inference 2003 - Object-Oriented Programming Model-Driven Programming “Data is as important as code” “Data is more important than code” Michael Daconta, Creating Relevance and Reuse with Targeted Semantics, XML 2004 Conference Keynote, November 16, 2004. 7

Formal Taxonomies as Ontologies Transportation Class Hierarchy OWL Listing: <? xml version="1. 0"? > <rdf: RDF xmlns: rdf="http: //www. w 3. org/1999/02/22 -rdfsyntax-ns#" xmlns: xsd="http: //www. w 3. org/2001/XMLSche ma#" xmlns: rdfs="http: //www. w 3. org/2000/01/rdfschema#" xmlns: owl="http: //www. w 3. org/2002/07/owl#" xmlns: daml="http: //www. daml. org/2001/03/daml +oil#" xmlns="http: //www. owlontologies. com/unnamed. owl#" xmlns: dc="http: //purl. org/dc/elements/1. 1/" xml: base="http: //www. owlontologies. com/unnamed. owl"> <owl: Ontology rdf: about=""/> <owl: Class rdf: ID="Transportation"/> <owl: Class rdf: ID="Air. Vehicle"> <rdfs: sub. Class. Of rdf: resource="#Transportation"/> </owl: Class> <owl: Class rdf: about="#Ground. Vehicle"> <rdfs: sub. Class. Of rdf: resource="#Transportation"/> </owl: Class> <owl: Class rdf: about="#Automobile"> <rdfs: sub. Class. Of> <owl: Class rdf: ID="Ground. Vehicle"/> </rdfs: sub. Class. Of> Etc. Formal Taxonomies for the U. S. Government, Michael Daconta, Metadata Program Manager, US Department of Homeland Security, XML. Com, http: //www. xml. com/pub/a/2005/01/26/formtax. html 8

Ontology-Driven Information Systems • Methodology Side – the adoption of a highly interdisciplinary approach (means multiple Co. Ps and effective coordination mechanisms): – Analyze the structure at a high level of generality. – Formulate a clear and rigorous vocabulary. • Architectural Side – the central role in the main components of an information system: – Information resources. – User interfaces. – Application programs. Source: Nicola Guarino, Formal Ontology and Information Systems, Proceedings of FOIS ’ 98, Trento, Italy, 6 -8 June 1998. 9

Categories of Ontologies Examples SUMO HL 7 RIM FEA-RMO* EON SNOMED CT LOINC Source: Netcentric Semantic Linking (Mapping): An Approach for Enterprise Semantic Interoperability, Mary Pulvermacher, et. Al. MITRE, October 2004. 10

Upper Ontologies – Several examples of Upper Ontologies illustrate their important functionality in practice; SUMO, DOLCE, Omega, MSO for example. – The Suggested Upper Merged Ontology (SUMO) and its domain ontologies form the largest formal public ontology in existence today. – They are being used for research and applications in search, linguistics and reasoning. SUMO is the only formal ontology that has been mapped to all of the Word. Net lexicon. – SUMO is written in the SUO-KIF language. – SUMO is free and owned by the IEEE. – The ontologies that extend SUMO are available under GNU General Public License. Adam Pease, a member of the Ontolog Forum, is the Technical Editor of SUMO. – See http: //www. ontologyportal. org/. 11

FEA Reference Model Ontology FEA-RMO* • The purpose of FEA-RMO is to: – Define an ontology based on FEA reference models (PRM, BRM, SRM, TRM, and DRM), – Develop a common vocabulary, or lexicon, from the FEA reference models, – Support execution, validation, and inference based on FEA reference models, – Support the GSA role in e-Government as market maker, and – Support OMB/AIC partnership in AIC Task 1 & AIC Task 4 by providing lessons learned an ontology. Source: GSA FEA Reference Model Ontology: A Domain Specific Parsimonious Ontology, 12 Rick Murphy, Enterprise Architect, Office of the CIO, GSA, January 18, 2005.

Business Reference Model Taxonomy • Four Business Areas-one of which is: – Services to Citizens, which has • 19 Lines of Business-one of which is: – Health, which has • 5 Topics: – – – Health Care Services Illness Prevention Immunization Management Public Health Monitoring Consumer Health and Safety "The Business Reference Model is a function-driven framework for describing the business operations of the Federal Government independent of the agencies that perform them. “ Federal Enterprise Architecture Program Management Office. 2004. 13

Health Domains Taxonomy • Access to Care – Focuses on the access to appropriate care • Population Health and Consumer Safety – Assesses health indicators and consumer products as a means to protect and promote the health of the general population • Health Care Administration – Assures that federal health care resources are expended effectively to ensure quality, safety, and efficiency • Health Care Delivery Services – Provides and supports the delivery of health care to its beneficiaries • Health Care Research and Practitioner Education – Fosters advancements in health discovery and knowledge Source: Introduction to the Federal Health Architecture Development Methodology, Briefing to the FHA APRG, February 10, 2005. 14

Health Domains Taxonomy • Health Care Domains • FHA Organization (simplified): (New): – – – Access Information Administration Delivery Services Research and Education – – – Regional Initiatives Clinical Practice Population Health Interoperability Federal Health Architecture Source: Architectural Peer Review Group (APRG) Initial Meeting, February 10, 2005 15

Demonstrators • Samson W. Tu, Stanford Medical Informatics, Stanford University: – – The EON and ATHENA Projects What the Clinician Sees Guideline-Based Decision Support Architecture The Ontology in Protégé • Peter Yim, Co-Organizer of the Ontolog Forum: – The SUMO Time-Related Concepts in Protégé – Upper Ontology Component Reuse: • Extend From an Upper Ontology (SUMO), Rather Than Repeating the Definitions Within the Domain Ontology (HL 7) 16

Demonstration #1 Health Care Domain • EON: A domain-independent, component-based architecture for automation of protocol-based care. • Architecture (see next slide): – Problem solving components that have task-specific functions: • Planning patient’s therapy • Determining patient’s eligibility for protocols – A temporal data mediator that • Extends the standard relational model with a model of time • Supports valid-time temporal queries and updates – A shared knowledgebase of protocols and general medical concepts Source: http: //www. smi. stanford. edu/projects/eon/96 SCAMCMusen/sld 007. htm 17

Demonstration: #1 Architecture Patient Database Protégé Knowledge Base Servers Temporal Mediator Protocol Eligibility Checker Medical Domain Ontology Guidelines Yenta Eligibility Yenta Client Therapy Advisory Server EON Guideline Ontology Patient Data Model Clients Advisory Yenta Client Protégé 18

Pause for Slides & Online Demo Samson’s 13 slides follow this slide 19

Reuse and Semantic Interoperability • Multiple working groups shouldn't redefine-basic concepts – Undermines semantic interoperability across domains and systems – Varying quality of individual models – Limits downstream extensibility • Ontology-based formalizations offer more rigor – – Typically leverage work of broader community of interests Designed for reuse and extensibility Generally reflect more thorough, higher-quality modeling Reuse of Upper and Mid-level ontologies improves semantic alignment of Domain-Level ontologies and resulting implementations 20

Examples of Time Formalization • HL-7* – Time taxonomy fragment – Timing. Event model • Word. Net – Time (Verb) – Time (Noun) • SUMO – Process Examples *See Patrick Cassidy notes: http: //ontolog. cim 3. net/forum/health-ont/2005 -02/msg 00011. html 21

HL-7 Timing Event Model Type Concept ID Mnemonic Description L: (AC) 10708 AC Before meal (from lat. ante cibus) L: (ACD) 10712 ACD Before lunch (from lat. ante cibus diurnus) L: (ACM) 10711 ACM Before breakfast (from lat. Ante cibus matutinus) L: (ACV) 10713 ACV Before dinner (from lat. Ante cibus vespertinus) L: (HS) 10707 HS The hour of sleep (e. b. , H 18 -22) L: (IC) 10710 IC Between meals (from lat. Inter cibus) L: (ICD) 10718 ICD Between lunch and dinner L: (ICM) 10717 ICM Between breakfast and lunch L: (ICV) 10719 ICV Between dinner and the hour or sleep L: (PC) 10709 PC After meal (from lat. post cibus) L: (PCD) 10715 PCD After lunch (from lat. post cibus diurnus) L: (PCM) 10714 PCM After breakfast (from lat. post cibus matutinus) L: (PCV) 10716 PCV After dinner (from lat. Post cibus vespertinus) 22

Time Representation in HL 7 The following was selected from the HL-7 taxonomy: • Data. Type. Data. Value • Data. Type. Interval – Data. Type. Interval. Of. Physical. Quantities – Data. Type. Interval. Of. Points. In. Time • Data. Type. Event. Related. Interval • Data. Type. General. Timing. Specification • Data. Type. Periodic. Interval. Of. Time • Data. Type. Quantity – Data. Type. Physical. Quantity • Data. Type. Parametric. Probability. Distribution. Of. Physical. Quantities – Data. Type. Point. In. Time 23

Word. Net “Time” (Verb) 1. 2. 3. 4. 5. S: (v) clock, time (measure the time or duration of an event or action or the person who performs an action in a certain period of time) "he clocked the runners" S: (v) time (assign a time for an activity or event) "The candidate carefully timed his appearance at the disaster scene" S: (v) time (set the speed, duration, or execution of) "we time the process to manufacture our cars very precisely" S: (v) time (regulate or set the time of) "time the clock" S: (v) time (adjust so that a force is applied an action occurs at the desired time) "The good player times his swing so as to hit the ball squarely“ From http: //wordnet. princeton. edu/ 24

Word. Net “Time” (Noun) 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. S: (n) time, clip (an instance or single occasion for some event) "this time he succeeded"; "he called four times"; "he could do ten at a clip" S: (n) time (an indefinite period (usually marked by specific attributes or activities)) "he waited a long time"; "the time of year for planting"; "he was a great actor is his time" S: (n) time (a period of time considered as a resource under your control and sufficient to accomplish something) "take time to smell the roses"; "I didn't have time to finish"; "it took more than half my time" S: (n) time (a suitable moment) "it is time to go" S: (n) time (the continuum of experience in which events pass from the future through the present to the past) S: (n) clock time, time (the time as given by a clock) "do you know what time it is? "; "the time is 10 o'clock" S: (n) fourth dimension, time (the fourth coordinate that is required (along with three spatial dimensions) to specify a physical event) S: (n) time (a person's experience on a particular occasion) "he had a time holding back the tears"; "they had a good time together" S: (n) meter, metre, time (rhythm as given by division into parts of equal duration) S: (n) prison term, sentence, time (the period of time a prisoner is imprisoned) "he served a prison term of 15 months"; "his sentence was 5 to 10 years"; "he is doing time in the county jail" 25

Time Representation in SUMO Word. Net N 1 maps to SUMO “Process” • Appearance as argument number 1 – documentation Process "Intuitively, the class of things that happen and have temporal parts or stages. Examples include extended events like a football match or a race, actions like Pursuing and Reading, and biological processes. The formal definition is: anything that lasts for a time but is not an Object. Note that a Process may have participants 'inside' it which are Objects, such as the players in a football match. In a 4 D ontology, a Process is something whose spatiotemporal extent is thought of as dividing into temporal stages roughly perpendicular to the time-axis. “ – subclass Process Physical From http: //sigma 2. cim 3. net: 8080/sigma/Browse. jsp? term=Process&kb=SUMO 26

SUMO “Process” Examples • Appearance as argument number 2 (1 of 5) – subclass Dual. Object. Process • dual object process is a subclass of process • Appearance as argument number 3 (1 of 27) – domain agent 1 Process • the number 1 argument of agent is an instance of process 27

SUMO “Process” Examples • Antecedent (1 of 4) (=> (and (instance ? PROC Process) (sub. Process ? SUBPROC ? PROC)) (exists (? TIME) (time ? SUBPROC ? TIME))) – if ? PROC is an instance of process and ? SUBPROC is a subprocess of ? PROC, then there exists ? TIME so that ? SUBPROC exists during ? TIME 28

Questions to consider • Ontolog group’s focus on interoperability needs of NHIN & FHA architects, – How should ontologies enable the interoperability of patient health records? – How should existing and prospective health domain ontologies and taxonomies be aligned with upper ontologies to improve the accuracy of conceptual information transfer ? – Especially among systems using different domain knowledge representations. 29

Next Steps for Federal Architects • Help with the Definitions and Relationships for the Health Domains – so they connect between the FEA Reference Model Ontology and Domain-Specific Ontologies. • Help with Reuse of Upper Ontology Components in Domain Ontologies. • Identify Other Examples of Domain Ontologies and Ontology-Driven Information Systems That Will Serve as Best Practice Examples. 30

31

32

33

34

Mark Musen, MD, Ph. D • Mark Musen, MD, Ph. D: – Professor of Medicine and, by courtesy, Computer Science, Stanford University School of Medicine. – Dr. Musen's research interests include • knowledge modeling in biology and medicine, knowledge management, • automated support for clinical-practice guidelines and for clinical trials, and • knowledge-based approaches to public-health surveillance. – http: //clinicalinformatics. stanford. edu/scci_seminars/2 004 -05. html 35