The Linked Data Value Chain Model: A Methodology for Information Integration and Orchestration

28  November  2013  

Daniel Hladky Semantic Web Lab at HSE/W3C

Na/onal  Research  University  Higher  School  of  Economics  

Background  Computer  Science  and  Economics  SAP  AG,  iXOS  (OpenText),  Ontos    

Research  Interest  Linked  (Open)  Data  for  Government  &  Enterprises  NLP,  Seman/c  Web,  Business  Impact  of  Linked  Data    Ac3vi3es  Researcher  EU  FP7  –  GeoKnow  (within  Ontos),  LOD  Russia  KESW  2012  (Lecture  Linked  Enterprise  Data),  KESW’13  (Co-­‐Chair)  PC  member  at  ISWC/WoLE  (2012,  2013),  MLW  Rome  (2013)  W3C  Russia  office  hosted  by  NRU  HSE  Moderator  Open  Data  at  RIAN    Co-­‐organising  Skolkovo  CREI  proposal  with  KAIST,  ICSI  Berkley,  EIS  Bonn      

NRU HSE / W3C Russia Slavyanskaya Sq. 4 Bldg. 2 109074 Moscow, Russia E: dhladky@hse.ru E: dhladky@w3c.org.ru http://www.hse.ru/org/hse/iit/semant/

About Me

2  

1. Introduction to Research project

3  

The Linked Data Value Chain Model: A Methodology for Information Integration and Orchestration

Finance Sales & Marketing

Equipment and assets

Finance Sales & Marketing

Equipment & Assets

Enterprise-Wide Reusable Information

4  

2. Topic in the research community

References (excerpt) •  D. Wood – Linking Enterprise Data •  T. Heath, C. Bizer – Linked Data… •  B. Hu, G. Svensson – A case study of

linked enterprise data •  FP. Servant – Linking Enterprise Data •  P. Frischmuth, S. Auer – Linked Data in

Enterprises Sources that talk about this are •  ISWC •  LDOW •  WWW •  Semantic-Web-Journal

EU funded projects Unstructured Data - Papers, Legacy documents E-mails, Blogs

Semi-Structured Data - Wikis, Calendars, CMS

Structured Data - ERP, CRM, RDBMS

Semantic Knowledge Base

Organization Data (Intranet) Public Data (External)

Fire

wal

l

Link

ed

Ope

n D

ata

New

s &

S

ocia

l Med

ia

4. Research Tasks and Objectives

5  

Harvest Normalize Semantize Enrich Build Expose

1.  Can we develop a simple and suitable model that captures the Linked Data paradigm in relation to the value chain?

2.  Does the Linked Data model support the understanding of information integration using Linked Data technologies?

3.  Can we define a model for measuring and quantifying the value of Linked Data information integration along the value chain?

4. Research Tasks and Objectives

6  

Criteria 1 “Model” A model to capture the Linked Data Value Chain. Criteria 2 “Value” Algorithm and method to measure the value by comparing the metrics from the manual and the tool supported process. Criteria 3 “Prototype” A software prototype that will demonstrate how the automation of the Linked Data stack can be orchestrated.

Information Integration ROI Valuation LD Orchestration

5. Importance

7  

Database Integration -  taxonomies, schemas

Fusion and Linking -  Across heterogeneous

data Web Services -  REST, SOA, API etc

Tools - MDM, BI, Portals

Ease of Use -  Simplifying process for

non expert users

Workbench / Platform -  Integration of tools Flexibility -  Configuration and

adoption to new needs

Justifying Investment -  Model to evaluate

investment

ROI Adoption -  Extend existing

models to the new Linked Data Paradigm

6. New Research Area

1. A New Model 1. Linked Data Value Chain Model 2. ROI algorithm to valuate the model

2. Linked Data Framework

1. Orchestrating the LD Process 2. Simplifying the information integration 3. Proof of Concept to measure impact

8  

6.1 Research Area – Linked Data Value Chain Model

En/ty  

Types of Data Linked Data Roles Participating Entities

Raw  Data  Provider  

Linked  Data  Provider  

Linked  Data  Applica/on  Provider  

End  User  

Raw  Data   Linked  Data   Human-­‐Readable  Data  

prov

ides

can act as can act as

9  

6.2 Linked Data Orchestration Framework

LDP  Orchestra3on  Framework  

Object  Link  Discovery  Ontology  mapper    Miner  /  MiniDix  

Ontology  to  RDB  mappings  

Ontologies   LOD  Discovery  Results  

Meta Information Database

OntoQuad  

RDF  Database  

Ontology  Ed

itor  

Ontology  to  RDB

 Map

per  

Instan

ce  Editor  

SPARQL  endpoint  

D2RQ

 /  R2R

ML   Automa3c  Link  

Discovery  

Manual  Edi3ng  and  Correc3on  

External  Rela3onal  Database  

WWW  

Download Ontologies

Create and modify Ontologies

Edit RDF Objects

Open Data Set - 2

Open Data Set - n

Open Data Set -1

Link RDF Database Objects to Open Data Set Objects

Map Ontology to RDB and generate RDF dataset

OntoQuad plays role of the central Storage for the collected triplified RDF data

10  

7. Experimental results

11  

7.1 Experimental results – Task comparison (simplified)

Task   Avg.  3me  before   Avg.  3me  with  LD  

12  

8. Comparison – ROI using the LDVCM

13  

9. Conclusions

Criteria  1  “Model”  A  new  Linked  Data  Value  Chain  Model  is  developed  and  captures  the  Linked  Data  Paradigm  for  Informa/on  Integra/on.  

Criteria  2  “Value”  The  LD  Value  Chain  ROI  Algorithm  is  suitable  to  measure  the  impact  and  efficiency.  A  significant  impact  is  shown  using  the  Linked  Data  Orchestra/on  Framework.      

Criteria  3  “Prototype”  A  Linked  Data  Orchestra/on  Framework  is  developed  and  tested  using  the  CRM  example,  LOD  Russia  and  GeoKnow.  

14  

10. Research Summary

Practical implementations

LOD Russia Ministry of Education and Science Russia, Contract № 07.524.11.4005 of October 20, 2011

GeoKnow – Generator EU FP7 funded project, Grant Agreement No.318159

DoW – Linked Data Orchestration Workbench Swiss CTI funded project

Papers (excerpt)  Hladky  D.,  Maltseva  S.V.,  “Linked  Data  Paradigm  for  Enterprises:  Informa/on  Integra/on  and  Value  Chain”,    Business  Informa/cs  Journal  No  24/2013    Hladky  D.,  Drobyazko  G.,  Klintsov  V.,  “Enabling  Russian  Na/onal  Knowledge  with  Linked  Open  Data“,  In  Proc.  of  ISWC/WoLE2012,  Boston,  11  November  2012    Hladky  D.,  “LOD  Russia  –  Enabling  Russian  Na/onal  Knowledge  with  scien/fic  Open  Data“,  In  Proc.  of  PMOD  –  Using  Open  Data,  Brussels,  Belgium,  10-­‐20  June  2012    Hladky  D.,  “Sustainable  Advantage  for  the  Investor  Rela/ons  Team  through  Seman/c  Content”,  Book  Chapter  for  “Seman/c  Web”,  ISBN  978-­‐953-­‐7619-­‐54-­‐1,  Publisher  In-­‐Tech,  DOI:  10.5772/7307,  2010    Efimenko  I.,  Hladky  D.,  Khoroshevsky  V.F.,  Klintsov  V.,  “Seman/c  Technologies  and  Informa/on  Integra/on:  Seman/c  Wine  in  Media  Wine-­‐skin”,  In  Proc.  of  the  2nd  European  Seman/c  Technology  Conference  (ESTC2008),  Vienna,  2008  

15  

Q&A  Thank  You  

16