INTEROPERABILITY IN THE EDUCATION SECTOR USING A SECTOR

INTER-OPERABILITY IN THE EDUCATION SECTOR USING A SECTOR DATA MODEL DRIVEN METHODOLOGY

Introduction • During 2008 SSC, Standards NZ and a number of agencies worked together on developing 3 sets of guidelines for governing XML inter-operability. • We recognised the need to share information on how different agencies are approaching the development and standardisation of XML messaging. • We wanted to include vendors, technical/IT people and those in the business of data management/business intelligence. Robyn Smits, Senior Manager, Sector Services, Ministry of Education

XML Seminar Agenda • Introduction • Background • Wider Context of XML Schema Design • Developing the Education Sector Data Model (ESDM) • From Data Model to XML Schemas • Lessons Learnt • Questions

Why go there? A business perspective Andreas Kubisch – Senior Business Systems Analyst

The journey so far Hand coded Schemas: • Early notification • as. TTle integration (EDUCTF) No Schema: • E-Returns ESDM derived Schemas: • ENROL Integration (3 message pairs) • SMS-LMS Integration (9 messages) • Student Record Transfer (6 messages) more to come …

The challenges • Code Value Lists • Versioning • Data quality • Message Validation • Connectivity • Vendor capability • Model flexibility

Wider Context of XML Schema Design • What is an XML Schema? Juerg Tschumperlin, DMS Data Management Solutions

Wider Context of XML Schema Design What is an XML Schema? • A schema is a way to describe and validate data in an XML environment. • A schema is a model for describing the structure of information. • XML Schema (XSD) is a recommendation of the W 3 C. Source: EPA: http: //www. epa. gov/webservices/glossary. htm

Wider Context of XML Schema Design What is an XML Schema? Example of XML Schema Diagram: (Extract)

Wider Context of XML Schema Design What is an XML Schema? Example of XML Schema Code: (Extract)

Wider Context of XML Schema Design What is an XML Instance Document? Example of XML Instance: (Extract)

Wider Context of XML Schema Design XML Instance: Validation against XML Schema Incoming Instance Validates? N = Return to Sender Y = Process Instance

Wider Context of XML Schema Design How are XML Schemas commonly developed? • By standard-setting bodies: “off-the-shelf” or • Custom-build to meet requirements.

Wider Context of XML Schema Design How are custom XML Schemas developed? Usually • Project-based, and Hand-coded OR • Project-based, and Generated from - database structure, or - application program code, or - XML Schema model (1 stand-alone model per schema)

Wider Context of XML Schema Design How are custom XML Schemas developed? Usually • Project-based, and Hand-coded Conventional schema development OR • Project-based, and Generated from - database structure, or - application program code, or - XML Schema model (1 stand-alone model per schema) creates serious issues: - XML tag name variations, and - data misinterpretation, and - data quality issues, and - schema look and feel variations, and - schema versioning variations, and - schema re-use variations - schema maintenance nightmare

Wider Context of XML Schema Design Fact: Most organisations having developed custom XML Schemas came to realise that; • Neither interoperability nor data quality has improved, • Master XML Schemas are scattered and hard to find, • Schema maintenance is unsustainable, • Unmanaged re-use / versioning is an operational risk.

Wider Context of XML Schema Design Facts: The XML Schema Language is not to blame. The real culprit is how we develop schemas.

Wider Context of XML Schema Design The challenge: How can we develop XML Schemas that are; • semantically consistent, • syntactically standardised, • consistent across projects and organisations, and hence improve; • semantic interoperability, • data quality, • operational robustness, • sustainable change management. ?

Wider Context of XML Schema Design Today’s presentation will show: Our early work in XML Schema development in the school sector: • Using a data model-driven approach, • Producing derived and generated XML Schemas, • Which adhere to, + the single messaging data model, and + several global and NZ e-Gif standards • With a centralised schema administration.

Wider Context of XML Schema Design The presented solution is also: A generic XML Schema development approach for creating standardised custom schemas: • Suitable for any industry, • Suitable for messaging between organisations. • Suitable for integrating in-house applications (EAI).

Wider Context of XML Schema Design Today’s presentation will show: ESDM Custom Data Model Education Sector Data Model feeds into MXV Transformation Model-driven XML Vocabulary Transformation Methodology derives XML Schemas

ESDM Project Education Sector Data Model: History and Outcomes Alistair Betts, Summit Solutions Ltd 22

Purpose of ESDM “To produce a sector data model that defines shared sector data so that sector participants can provide, manage, access and understand the data. ” 23

Scope • Logical Model of Shared Sector Data • Data Custodianship responsibilities • Methodology for use • Report of Findings • Review of Data Modelling tools (Ministry of Education only) 24

The Model Subject Areas 25

Agencies Consulted • Career Services • Teachers Council • ERO • NZQA • TEC • MOE 26

Creating a message How do we go about creating a message using this methodology? 1. Is it a sector message? 2. Map requirements to model & mock up messages 3. Approve changes to ESDM via data dictionary 4. Update ESDM model 5. Create new draft message schemas 6. Model & message approval 7. Final changes and release to specification status 27

From Data Model to XML Schema Juerg Tschumperlin, D-M-S Data Management Solutions

From Data Model to XML Schema Industry-independent overview: Custom Data Model feeds into MXV Transformation Methodology derives XML Schemas

From Data Model to XML Schema Education Sector overview: ESDM Custom Data Model Education Sector Data Model feeds into MXV Transformation Methodology derives ESL XML Schemas Education Sector Language

From Data Model to XML Schema g n i l Data Owners d e n d a o t s Administrator m r e e l E d g n n n ESDM u Datad. Model i Custom Schema s Education Sector Datato Model a t a -en Consumers a n d i dt d d oo e e r MXV Transformation t l a n su h feeds into Methodology e Project S Team m pp u or c t o D ESL XML Schemas Main players Messaging Requirements Tools Project Team derives XML Schemas Education Sector Language

a From Data Model to XML Schema L ESDM ed s i d r a d M X d n a de m e ch S ti i fin s n o p nis UTF-8 character encoding mrga o c o ISO 11179 -5 Data Naming if nd es m ng el -Gs a n a a n ni fe e ct t g. Transformation SMXV o d Z oje i a s n. N r r a - t Methodology e k h sp v - looig os - hacr - Education Sector Data Model es c n n o i lia at W 3 C XML Schema W 3 C XML OASIS Universal Business Language Naming and Design Rules (UBL NDR 2) [adapted] UN/CEFACT Unqualified Data Types Schema feeds into URN Naming derives ESL XML Schemas Education Sector Language

From Data Model to XML Schema : s a c m n e y h s c ESDM S in Education Sector Data Model L L e M M X X ch cad d t d s e a n ir v a m ca L s ns e D UM e io - Nam init f XML Schemas e ESL equivalent expressions D - Education Sector Language

From Data Model to XML Schema ESDM Education Sector Data Model MXV Transformation Methodology populates XML Component Library equivalent expressions is referenced by ESL XML Schemas Education Sector Language

From Data Model to XML Schema UML data model XML Schema Component Library Modules: Domain / Data Type - UDT UN/CEFACT Unqualified Data Types (read-only) Attribute ESDM Education Sector Data Model OASIS UB - QDT Qualified Data Types - CBC Common Basic Components L NDR-lik e compon XML Component library Re-use ofpopulates XML Sche. Library ma comp onents Tool supp orted Imp XML Schemas aiscreferenced t Anabylys. ESL Sector Language i. Education s Class Specialization Association - CAC Common Aggregate Components MXV Transformation Methodology

From Data Model to XML Schema ESDM Education Sector Data Model feeds into MXV Transformation Methodology derives ESL XML Schemas Education Sector Language

From Data Model to XML Schema MXV Transformation Methodology

From Data Model to XML Schema MXV Transformation Methodology Concepts Standards Tools UML Modelling ISO 11179 -5 Data Naming Change Management Process Model-driven Architecture Semantic Vocabulary (aka Ontology) W 3 C XML Schema W 3 C XML OASIS Universal Business Language Naming and Design Rules (UBL NDR 2) [adapted] UN/CEFACT Unqualified Data Types Schema Sybase Power. Designer D-M-S MXV Productivity Tools (optional)

ds r a d From Data Model to XML Schema n a t s L M X c i t Data Owners man se m o I) t Administrator s A u E c ( s g s n n n i o o i p i t t ESDM lo Model a a Schema c e s Custom Data i i l v n p Educationd. Sector Data Model Transformation Methodology e. MXV a p Consumers g a r r o o f Concepts Standards se Tools n y n u r e o t o i e UML Modelling t s h u u w t n i ol e nd s i b n Data Naming ISO 11179 -5 e e y g l e in W 3 C XMLt. Schema an. Architecture w feeds into g tr ab Model-driven W 3 C XML Sybase Power. Designer r a e o o s f b s P e g Universal Business Language Naming and e l. Semantic Vocabulary OASIS D-M-S MXV Productivity Tools n b i m a g (aka Ontology) Design Rules (UBL NDR 2) [adapted] (optional) t r i a o u s f s S e e l UN/CEFACT Unqualified Data Types Schema b m a ESL XML Schemas it or XML Schemas u f generates S Education Sector Language e l b a t i Su Change Management Process

From Data Model to XML Schema MXV Tool Support Administrator UML Modelling (PD) Custom Data Model Create New Model Version (DMS) Check NDR Naming Reand Design Rules Compliance (DMS) Automatically populatelia the XML Component Library (DMS) b l Fa of XMLe Component sc Check completeness Library (DMS) st MXV Transformation h em (PD) tur Analysis UMLfeeds to XMLinto Schema Impact n-a aq Methodology rou Modelling ua (PD) XML Schema lity nd Migrate to Specification tim (DMS) e Clone an XML View (DMS) MXV Tools: PD = Sybase Power. Designer DMS = MXV Productivity Tools XML Schema (PD & DMS) XMLGenerator Schemas generates Create Delivery Package (DMS)

From Data Model to XML Schema Further Opportunities 1) Extensions (optional) Extension Organisation A Custom Data Model Shared Data of Organisation A, B and C Extension Organisation C e l b XML Component i s n Library e t x E MXV Transformation Methodology feeds into generates Extension Organisation A Extension Organisation C XML Schemas In-house Schema of Organisation A

From Data Model to XML Schema Further Opportunities: 2) Value Validation (optional) A standard methodology for validating values in an XML instance • Using OASIS Genericode format for defining code value lists • Using OASIS Context Value Association format for associating XML schemas to code value lists • Using ISO Schematron format for defining more complex business constraints • Generates one run-time Stylesheet for platform-independent validation

From Data Model to XML Schema Further Opportunities: 2) Value Validation (optional 2 -pass) Diagram courtesy of www. cranesoftwrights. com

From Data Model to XML Schema Further Opportunities: 2) Value Validation (optional) Custom Data Model feeds into Tools i l a a d n o ti V MXV Transformation e u l Methodology a V d XML Value e generates t a Validation skeletons r g egenerates XML Schemas t In

From Data Model to XML Schema Further information available Brochure • Model-driven Semantic Interoperability: Education Sector Case Study Demo CD • Model-driven XML Vocabulary: 1 hr demo video showing MXV in action SSC Object Registry Service • http: //registry. e. govt. nz Web links • http: //www. d-m-s. co. nz/serv_xmlschema. htm (incl. MXV Trial Download)

Credits and References Standards UML Unified Modelling Language http: //www. uml. org/ W 3 C XML 1. 0 http: //www. w 3. org/TR/1998/REC-xml-19980210 W 3 C XML Schema http: //www. w 3. org/XML/Schema#dev ISO/IEC 11179 -5 Part 4 Data Element Naming http: //www. iso. org/iso_catalogue/catalogue_tc/catalogue_detail. htm? csnumber=35347 UN/CEFACT Unqualified Data Types Schema http: //www. unece. org/cefact/ OASIS UBL NDR http: //docs. oasis-open. org/ubl/os-UBL-2. 0/UBL-2. 0. html New Zealand Education Sector Language http: //www. minedu. govt. nz/~/media/Min. Edu/Files/Education. Sectors/Primary. Secondary/Initiatives/Model. Driven. Semantic. In eroperability. pdf OASIS Genericode and OASIS Context Value Association http: //www. oasis-open. org/committees/tc_home. php? wg_abbrev=codelist http: //www. genericode. org ISO Schematron http: //standards. iso. org/ittf/Publicly. Available. Standards/index. html W 3 C XSLT http: //www. w 3. org/TR/xslt 20/ Tools Modelling Sybase Power. Designer http: //www. sybase. com/products/modelingdevelopment/powerdesigner XML Value Validation Crane Softwrights Schematron Implementation of OASIS context/value association files http: //www. cranesoftwrights. com Crane Softwrights Genericode and context/value association files rendering stylesheets http: //www. cranesoftwrights. com ISO Schematron to XSLT skeleton http: //www. schematron. com

Lessons Learnt Robyn Smits, Ministry of Education

Lesson’s learnt • Keys to success: – Well defined message requirements assessed early in project lifecycle – Message mock-ups presented to subject matter experts during requirements definition phase – Early confirmation of modeller interpretations & issues – Similarity of messages reduced time required to create subsequent messages after first created – Stability of requirements meant little rework – Good project communications. Engage your data modeler early 48

Lessons Learnt • • • • Accept interoperability as an ongoing subject matter. Ensure full business and IT management support. Allow for broad semantic model consensus. Expect consensus to be time consuming. Ensure long-term semantic model ownership. Evolving the data model keeps schema forward migrations simple. Avoid data model revolutions if possible. Keep administration centralised. Ascertain strong tool support before embarking. Except impact on business processes and processing. Expect impact on application development and testing. Interoperability attempts without IT governance are likely to fail. Plan for a cultural change. Don’t expect efficiency gains too soon. Manage the optimal component library release frequency.

Questions ?

The verdict It delivers: • Standardised messages, • Powerful validation options. But it requires: • Requires whole organisation commitment, and • Schema creation is complex.

Feel free to contact us • Robyn Smits, Mo. E r u o y r o f u o y k n a d Th Andreas Kubisch, Mo. E n a time e c n a d n e t at Alistair Betts, Summit Solutions Senior Manager Sector Services www. minedu. govt. nz • Senior Business Systems Analyst, Interoperability Team www. minedu. govt. nz • ESDM Data Modeller, ESL Schema Developer www. summitsolutions. co. nz • Juerg Tschumperlin ESL / MXV Methodology, MXV Tools, Customisation and Implementation www. d-m-s. co. nz