D Frankel Semantic Interoperability v01-01 · Agenda • The$Need$for$Smarter$Systems$ •...

27
The Seman)c Interoperability Problem Business Impact and Technical Solu)ons A Pragma)c Approach to Improving Data Integra)on David Frankel Consul)ng March, 2013

Transcript of D Frankel Semantic Interoperability v01-01 · Agenda • The$Need$for$Smarter$Systems$ •...

Page 1: D Frankel Semantic Interoperability v01-01 · Agenda • The$Need$for$Smarter$Systems$ • A$technique$for$Mi)gang$the$Problem$ • Applicaon$of$the$Techniques$ 3/19/13 ©David$Frankel$Consul)ng$2013$

The  Seman)c  Interoperability  Problem  Business  Impact  and  Technical  Solu)ons  

A  Pragma)c  Approach  to    Improving  Data  Integra)on  

 David  Frankel  Consul)ng  

March,  2013  

Page 2: D Frankel Semantic Interoperability v01-01 · Agenda • The$Need$for$Smarter$Systems$ • A$technique$for$Mi)gang$the$Problem$ • Applicaon$of$the$Techniques$ 3/19/13 ©David$Frankel$Consul)ng$2013$

Agenda  

•  The  Need  for  Smarter  Systems  •  A  technique  for  Mi)ga)ng  the  Problem  •  Applica)on  of  the  Techniques  

3/19/13   ©David  Frankel  Consul)ng  2013   2  

Page 3: D Frankel Semantic Interoperability v01-01 · Agenda • The$Need$for$Smarter$Systems$ • A$technique$for$Mi)gang$the$Problem$ • Applicaon$of$the$Techniques$ 3/19/13 ©David$Frankel$Consul)ng$2013$

Agenda  

•  The  Need  for  Smarter  Systems  •  A  technique  for  Mi)ga)ng  the  Problem  •  Applica)on  of  the  Techniques  

3/19/13   ©David  Frankel  Consul)ng  2013   3  

Page 4: D Frankel Semantic Interoperability v01-01 · Agenda • The$Need$for$Smarter$Systems$ • A$technique$for$Mi)gang$the$Problem$ • Applicaon$of$the$Techniques$ 3/19/13 ©David$Frankel$Consul)ng$2013$

The  Seman)c  Interoperability  Barrier  A  Business  View  

•  Seman)c  interoperability  of  data:  The  ability  of  mul/ple  par/es  to  coordinate  their  func/on  based  on  a  shared  understanding  of  the  data  that  flows  among  them  

•  Business  Drivers  –  Value  chains,  outsourcing,  M&A,  industrial  convergence,  need  for  transparency  e.g.  for  regula)on  and  risk  management  

•  Data  integra)on  boQlenecks  measurably  impact  global  GDP  – Modest  improvements  have  large  economic  ramifica)ons  

3/19/13   ©David  Frankel  Consul)ng  2013   4  

Page 5: D Frankel Semantic Interoperability v01-01 · Agenda • The$Need$for$Smarter$Systems$ • A$technique$for$Mi)gang$the$Problem$ • Applicaon$of$the$Techniques$ 3/19/13 ©David$Frankel$Consul)ng$2013$

Seman)c  Interoperability    Technical  Factors  

•  Legacy  data  used  for  unintended  purposes  •  Standards  silos  

– E.g.  different  standards  with  different  architectures  for  trading  front  office  and  back  office,  electronic  submission  of  financial  statements,  payment  ini)a)on,  etc.  

– All  of  which  have  to  integrate  with  tradi)onal  rela)onal  databases  

•  New  data  sources  :  mobile  devices,  sensors    3/19/13   ©David  Frankel  Consul)ng  2013   5  

Page 6: D Frankel Semantic Interoperability v01-01 · Agenda • The$Need$for$Smarter$Systems$ • A$technique$for$Mi)gang$the$Problem$ • Applicaon$of$the$Techniques$ 3/19/13 ©David$Frankel$Consul)ng$2013$

Legacy  Gets  Stretched  

3/19/13   ©David  Frankel  Consul)ng  2013   6  

This graphic courtesy of Visa Inc.

Data Content Requirements Expand  

Product Processor Banks Category Merchant Channel

Data Design circa 1980

BIN MCC Terminal Names

Page 7: D Frankel Semantic Interoperability v01-01 · Agenda • The$Need$for$Smarter$Systems$ • A$technique$for$Mi)gang$the$Problem$ • Applicaon$of$the$Techniques$ 3/19/13 ©David$Frankel$Consul)ng$2013$

Standards  Silos  

3/19/13   ©David  Frankel  Consul)ng  2013   7  

ISO  20222   FpML   FIX   TWIST   XBRL   BIAN  

Page 8: D Frankel Semantic Interoperability v01-01 · Agenda • The$Need$for$Smarter$Systems$ • A$technique$for$Mi)gang$the$Problem$ • Applicaon$of$the$Techniques$ 3/19/13 ©David$Frankel$Consul)ng$2013$

Today’s  Model-­‐Driven    Data  Integra)on  Tools  

3/19/13   ©David  Frankel  Consul)ng  2013   8  

Field  1  

Coupon  Value  

Field  3  

Field  4  

Field  5  

Field  6  

Field  1  

Field  2  

Coupon  Value  

Field  4  

Field  5  

Field  6  

Expression:   Format  2  Field  4  =  Format  1  Field  5  *  Format  2  Field  6  

Page 9: D Frankel Semantic Interoperability v01-01 · Agenda • The$Need$for$Smarter$Systems$ • A$technique$for$Mi)gang$the$Problem$ • Applicaon$of$the$Techniques$ 3/19/13 ©David$Frankel$Consul)ng$2013$

What  We  Have  Achieved  to  Date  With  Model-­‐Driven  Integra)on  Tools  

•  More  rapid  route  to  executable  transforma)on  – ~20%  improvement  – Modest  improvement  à  Big  economic  return  

•  But  only  aier  analyst  figures  out  what  maps  to  what  – Figuring  that  out  is  the  remaining  80%  of  the  boQleneck  

3/19/13   ©David  Frankel  Consul)ng  2013   9  

Page 10: D Frankel Semantic Interoperability v01-01 · Agenda • The$Need$for$Smarter$Systems$ • A$technique$for$Mi)gang$the$Problem$ • Applicaon$of$the$Techniques$ 3/19/13 ©David$Frankel$Consul)ng$2013$

Another  Challenge  Finding  Service  Agents  

•  Many  service  agents  from  different  sources  •  Example:  BIAN  External  Liquidity  Risk  Scenario  

3/19/13   ©David  Frankel  Consul)ng  2013   10  

Graphic  created  by  James  Odell  

Page 11: D Frankel Semantic Interoperability v01-01 · Agenda • The$Need$for$Smarter$Systems$ • A$technique$for$Mi)gang$the$Problem$ • Applicaon$of$the$Techniques$ 3/19/13 ©David$Frankel$Consul)ng$2013$

Agenda  

•  The  Need  for  Smarter  Systems  •  A  technique  for  Mi)ga)ng  the  Problem  •  Applica)on  of  the  Techniques  

3/19/13   ©David  Frankel  Consul)ng  2013   11  

Page 12: D Frankel Semantic Interoperability v01-01 · Agenda • The$Need$for$Smarter$Systems$ • A$technique$for$Mi)gang$the$Problem$ • Applicaon$of$the$Techniques$ 3/19/13 ©David$Frankel$Consul)ng$2013$

Improving  Seman)c  Interoperability  

•  Leveraging  Seman)c  Technology  •  Chipping  Away  at  a  Large  Problem  •  Remember:    

– Modest  improvement  à  big  impact  

•  Key  challenge:  Techniques  must  be  non-­‐disrup/ve  – Cannot  require  ISO  20022,  XBRL,  BIAN,  Rela)onal  models  to  change  their  structure  

3/19/13   ©David  Frankel  Consul)ng  2013   12  

Page 13: D Frankel Semantic Interoperability v01-01 · Agenda • The$Need$for$Smarter$Systems$ • A$technique$for$Mi)gang$the$Problem$ • Applicaon$of$the$Techniques$ 3/19/13 ©David$Frankel$Consul)ng$2013$

Integra)on  via  Common  Seman)c  Anchoring  Mechanism  

3/19/13   ©David  Frankel  Consul)ng  2013   13  

ISO  20222  Messages  

XBRL  Reports  

CCTS  Models  

BIAN  Services  

Amor)zed  

Roll  Over  

…  

Business  Vocabularies/Ontologies  (Federated)    

An  addi)onal,  non-­‐disrup)ve  dimension  

Page 14: D Frankel Semantic Interoperability v01-01 · Agenda • The$Need$for$Smarter$Systems$ • A$technique$for$Mi)gang$the$Problem$ • Applicaon$of$the$Techniques$ 3/19/13 ©David$Frankel$Consul)ng$2013$

Making  Seman)c  Vocabularies  and  Ontologies  Ac/onable  

3/19/13   ©David  Frankel  Consul)ng  2013   14  

Asset_RollOver  …  Amor)zed_Assessment  :  Instrument_Amt  FairValue_Assessment  :  Instrument_Amt  

What  is  the  connec)on?  

IT  Data  Structure  (Rela)onal,  ISO  20022,  XBRL,  etc.)  

Business  Vocabulary/Ontology    (e.g.  FIBO)  

   

Amor)zed  

Roll  Over  

…  

Asset  Roll  Over  

Page 15: D Frankel Semantic Interoperability v01-01 · Agenda • The$Need$for$Smarter$Systems$ • A$technique$for$Mi)gang$the$Problem$ • Applicaon$of$the$Techniques$ 3/19/13 ©David$Frankel$Consul)ng$2013$

Seman)c  Structure  Based  on  ISO  11179  and  UN/CEFACT  

3/19/13   ©David  Frankel  Consul)ng  2013   15  

Field   Object   Property   Data  Type  

Amor6zed_Assessment   Asset  Roll  Over   Amor)zed  Assessment   Instrument  Amount  

FairValue_Assessment   Asset  Roll  Over   Fair  Value  Assessment   Instrument  Amount  

…  

Business  Vocabulary/Ontology  Fair  Value  Assessment  

Asset  Roll  Over  

…  

URI  

Page 16: D Frankel Semantic Interoperability v01-01 · Agenda • The$Need$for$Smarter$Systems$ • A$technique$for$Mi)gang$the$Problem$ • Applicaon$of$the$Techniques$ 3/19/13 ©David$Frankel$Consul)ng$2013$

Refined  Seman)c  Structure  

3/19/13   ©David  Frankel  Consul)ng  2013   16  

Field   Object  Qualifier  Term  

Object    Term  

Property  Qualifier  Term    

Property  Term  

DataType  Qualifier  Term  

CoreDataType  Term  

Amor6zed_Assessment  

Asset   Roll  Over  

Amor)zed   Assessment   Instrument   (Money)  Amount  

FairValue_Assessment  

Asset   Roll  Over  

Fair  Value   Assessment   Instrument   (Money)  Amount  

…  

Business  Vocabulary/Ontology  

Fair  Value  

Roll  Over  

…  

URI  

Page 17: D Frankel Semantic Interoperability v01-01 · Agenda • The$Need$for$Smarter$Systems$ • A$technique$for$Mi)gang$the$Problem$ • Applicaon$of$the$Techniques$ 3/19/13 ©David$Frankel$Consul)ng$2013$

Seman)c  Structure  of  a  Service  

3/19/13   ©David  Frankel  Consul)ng  2013   17  

Service  Opera6on   Ac6on   Object   Property  

RollOver_Asset_Balance   Ini)ate   Asset  Roll  Over  

…  

Amor)zed  

Asset  Roll  Over  

…  Business  Vocabulary/Ontology  

URI  

Page 18: D Frankel Semantic Interoperability v01-01 · Agenda • The$Need$for$Smarter$Systems$ • A$technique$for$Mi)gang$the$Problem$ • Applicaon$of$the$Techniques$ 3/19/13 ©David$Frankel$Consul)ng$2013$

Business  Context  Metadata  •  Geopoli)cal  scope:  The  con)nent,  economic  region,  country,  or  region  for  which  

the  element  is  relevant  •  Industry  classifica)on:  The  industry  or  industries  for  which  the  element  is  relevant  •  Product  classificia)on:  The  product  class(es)  for  which  the  element  is  relevant  •  Official  constraint:  A  legal  requirement(s)  associated  with  the  element;  e.g.,  an  

element  could  be  required  by  Sarbanes-­‐Oxley    •  System  capability:    Indicates  that  the  element  is  there  in  order  to  address  the  

limita)ons  of  a  compu)ng  system(s)  •  Business  process:  The  business  process(es)  with  which  the  element  is  associated  •  Business  process  role:  The  role(s)  directly  involved  in  the  business  process(es)  that  

are  relevant  to  the  element,  e.g.  shipper  •  Suppor)ng  role:  The  role(s)  indirectly  involved  in  the  business  process;  e.g.,  a  data  

element  in  an  order  response  from  seller  to  buyer  could  be  required  by  a  third-­‐party  shipper    

•  …  

3/19/13   ©David  Frankel  Consul)ng  2013   18  

Page 19: D Frankel Semantic Interoperability v01-01 · Agenda • The$Need$for$Smarter$Systems$ • A$technique$for$Mi)gang$the$Problem$ • Applicaon$of$the$Techniques$ 3/19/13 ©David$Frankel$Consul)ng$2013$

Seman)c  Structure  Metadata  +  Business  Context  Metadata  

3/19/13   ©David  Frankel  Consul)ng  2013   19  

Amor)zed_Assessment  :  Instrument_Amount  

Official Constraint: IFRS Fair Value Measurement

Property Qualifier

CoreDataType Term

FairValue_Assessment    :  Instrument_Amount  

Property Term

Official Constraint: US GAAP FASB Update No. 2011-04

DataType Qualifier

Page 20: D Frankel Semantic Interoperability v01-01 · Agenda • The$Need$for$Smarter$Systems$ • A$technique$for$Mi)gang$the$Problem$ • Applicaon$of$the$Techniques$ 3/19/13 ©David$Frankel$Consul)ng$2013$

Modeling  Seman)c  Structure  Addi)onal  Benefit  

Interpreta)on  #  1  •  Ac)on:  Create  •  Object:  Current  Account    •  Object  Qualifier:  New  •  Property:  Inventory  

Interpreta)on  #  2  •  Ac)on:  Create  •  Object:  Current  Account  •  Property:  Inventory  •  Property  Qualifier:  New  

3/19/13   ©David  Frankel  Consul)ng  2013   20  

Service:  Create  New  Current  Account  Inventory    

Page 21: D Frankel Semantic Interoperability v01-01 · Agenda • The$Need$for$Smarter$Systems$ • A$technique$for$Mi)gang$the$Problem$ • Applicaon$of$the$Techniques$ 3/19/13 ©David$Frankel$Consul)ng$2013$

Agenda  

•  The  Need  for  Smarter  Systems  •  A  technique  for  Mi)ga)ng  the  Problem  •  Applica)on  of  the  Techniques  

3/19/13   ©David  Frankel  Consul)ng  2013   21  

Page 22: D Frankel Semantic Interoperability v01-01 · Agenda • The$Need$for$Smarter$Systems$ • A$technique$for$Mi)gang$the$Problem$ • Applicaon$of$the$Techniques$ 3/19/13 ©David$Frankel$Consul)ng$2013$

Toward  a  New  Genera)on  of    Data  Integra)on  Tools  

3/19/13   ©David  Frankel  Consul)ng  2013   22  

Field  1  

Coupon  Value  

Field  3  

Field  4  

Field  5  

Field  6  

Field  1  

Field  2  

Coupon  Value  

Field  4  

Field  5  

Field  6  

                                         

Object  Class  Term:  Order  Object  Class  Qualifier:  Sales  Property  Term:  Coupon  CoreDataType  Term:  Amount  Product  Classifica)on:  Consumer  

Object  Class  Term:  Order  Object  Class  Qualifier:  Sales  Property  Term:  Coupon  CoreDataType  Term:  Amount  Product  Classifica)on:  Financial  

Coupon  

Coupon  

Page 23: D Frankel Semantic Interoperability v01-01 · Agenda • The$Need$for$Smarter$Systems$ • A$technique$for$Mi)gang$the$Problem$ • Applicaon$of$the$Techniques$ 3/19/13 ©David$Frankel$Consul)ng$2013$

Integra)on  via  Common  Seman)c  Anchoring  Mechanism  

3/19/13   ©David  Frankel  Consul)ng  2013   23  

ISO  20222  Messages  

XBRL  Reports  

CCTS  Models  

BIAN  Services  

Amor)zed  

Roll  Over  

…  

Business  Vocabularies/Ontologies  (Federated)    

An  addi)onal,  non-­‐disrup)ve  dimension  

Page 24: D Frankel Semantic Interoperability v01-01 · Agenda • The$Need$for$Smarter$Systems$ • A$technique$for$Mi)gang$the$Problem$ • Applicaon$of$the$Techniques$ 3/19/13 ©David$Frankel$Consul)ng$2013$

Conclusion  •  Large  problem  with  business  and  technical  causes  •  New  techniques  promise  modest  yet  significant  return  

–  Not  the  only  angle  of  aQack  •  Implementa)ons  under  way    

–  BIAN  and  CCTS  most  advanced  –  ISO  20022  and  XBRL  work  in  progress  

•  Applica)on  to  LEIs  and  standardized  instrument  IDs  –  A  new  CoreDataType:  LEI  ?  –  LEIs  for  product  classifica)on  and  other  business  context  metadata  elements  ?  

–  Further  discussion  to  ensue…    

 3/19/13   ©David  Frankel  Consul)ng  2013   24  

Page 25: D Frankel Semantic Interoperability v01-01 · Agenda • The$Need$for$Smarter$Systems$ • A$technique$for$Mi)gang$the$Problem$ • Applicaon$of$the$Techniques$ 3/19/13 ©David$Frankel$Consul)ng$2013$

For  More  Informa)on  

www.DFrankelConsul)ng.com/ar)cles.html    

Or  contact:    David  S.  Frankel  Email:  david@DFrankelConsul)ng.com  Mobile:  +1  530-­‐519-­‐4135  

3/19/13   ©David  Frankel  Consul)ng  2013   25  

Page 26: D Frankel Semantic Interoperability v01-01 · Agenda • The$Need$for$Smarter$Systems$ • A$technique$for$Mi)gang$the$Problem$ • Applicaon$of$the$Techniques$ 3/19/13 ©David$Frankel$Consul)ng$2013$

Backup  The  22  UN/CEFACT  Core  Data  Types  

   •  (Money)  Amount  •  Binary  Object  •  Code  •  Date  •  Date  Time  •  Dura)on  •  Graphic  •  Iden)fier  •  Indicator  •  Measure  •  Name  

   •  Ordinal  •  Percent  •  Picture  •  Quan)ty  •  Rate  •  Ra)o  •  Sound  •  Text  •  Time  •  Value  •  Video  

3/19/13   ©David  Frankel  Consul)ng  2013   26  

Page 27: D Frankel Semantic Interoperability v01-01 · Agenda • The$Need$for$Smarter$Systems$ • A$technique$for$Mi)gang$the$Problem$ • Applicaon$of$the$Techniques$ 3/19/13 ©David$Frankel$Consul)ng$2013$

WordNet  Copyright  Slide  number  16  quotes  defini)ons  from  WordNet,  the  online  dic)onary  project  of  Princeton  University.    WordNet  is  freely  available  for  this  kind  of  reuse,  as  long  as  it  is  accompanied  by  a  required  copyright  no)ce  (see  hQp://wordnet.princeton.edu/wordnet/license/).    Here  is  that  copyright  no)ce:    WordNet  3.0  Copyright  2006  by  Princeton  University.  All  rights  reserved.  THIS  SOFTWARE  AND  DATABASE  IS  PROVIDED  "AS  IS"  AND  PRINCETON  UNIVERSITY  MAKES  NO  REPRESENTATIONS  OR  WARRANTIES,  EXPRESS  OR  IMPLIED.  BY  WAY  OF  EXAMPLE,  BUT  NOT  LIMITATION,  PRINCETON  UNIVERSITY  MAKES  NO  REPRESENTATIONS  OR  WARRANTIES  OF  MERCHANT-­‐  ABILITY  OR  FITNESS  FOR  ANY  PARTICULAR  PURPOSE  OR  THAT  THE  USE  OF  THE  LICENSED  SOFTWARE,  DATABASE  OR  DOCUMENTATION  WILL  NOT  INFRINGE  ANY  THIRD  PARTY  PATENTS,  COPYRIGHTS,  TRADEMARKS  OR  OTHER  RIGHTS.    WordNet  is  a  registered  trademark  of  Princeton  Univiersity.    

3/19/13   ©David  Frankel  Consul)ng  2013   27