Session ID: MDM202 Configuration of Matching-Strategies For SAP Master Data Management

Session ID: MDM202 Configuration of Matching-Strategies for SAP Master Data Management
Contributing Speaker(s)
Christian Behre
NetWeaver Product Management, SAP AG
Remo Durante
Solution Architect, SAP Deutschland AG & Co.KG
SAP AG 2004, SAP TechEd / Session MDM202 / 2
Learning Objectives
As a result of this workshop, you will be able to:

understand the role of the Master Data Management understand the architecture of the Content Integrator understand the matching process explain the Content Integrator settings understand the meaning of Normalization and Matching Algorithm and how to adopt them
Overview Architecture Normalization Matching Algorithm Implementation Considerations
MDM A Key Capability of SAP NetWeaver 04

SAP NETWEAVER ONE PLATFORM ONE PRODUCT
SAP NetWeaver
PEOPLE INTEGRATION
SAP NetWeaver MDM

Enables companies to store,
augment, and consolidate

master data with high quality Ensures consistent distribution to all applications and systems within a
Multi channel access Composite Application Framework Portal Collaboration
INFORMATION INTEGRATION
heterogeneous IT landscape
Life Cycle Mgmt
Bus. Intelligence Master Data Mgmt
Knowledge Mgmt
Leverages existing IT investments in business-critical data Delivers vastly reduced TCO by effective master data management ensuring cross-system data consistency Accelerates and improves the
PROCESS INTEGRATION
Integration Broker
Business Process Mgmt
APPLICATION PLATFORM
J2EE
ABAP
DB and OS Abstraction
execution of business processes
SAP NetWeaver: Delivering on the promise of ESA

SAP NetWeaver
PEOPLE INTEGRATION Multi channel access COMPOSITE APPLICATION FRAMEWORK Portal Collaboration
Procurement Sales Shipment
INFORMATION INTEGRATION Knowledge Mgmt Bus. Intelligence
Master Data Management Master Data Management PROCESS INTEGRATION Integration Broker Bus. Process Mgmt
APPLICATION PLATFORM J2EE ABAP
DB and OS Abstraction
R/3
SRM
Siebel
i2
Legacy
...
Process Overview of SAP MDM Loading Master Data

SAP NetWeaver
Enterprise Portal Master Data Management
1
II 1
BI
Staging
4 4
4 1
4 Knowledge Mgmt.
Exchange Infrastructure
3 3
ERP Application Platform

I
CRM
Legacy
3rd Party
Legacy
Loading Master Data with Extraction MDM triggers the load PULL mechanism
II
Loading Master Data with Periodic Inbound Collector Client triggers the load PUSH principle
Process Overview of SAP MDM Consolidating Master Data

SAP NetWeaver
CI 5 6
? =
BI
Staging
1b 1a 2
?
4
Database
7 Knowledge Mgmt.
=

1a Consolidating Master Data from Extraction 1b
CRM
Legacy
3rd Party
Legacy
Consolidating Master Data from Staging (PIC)
Valid for both process variants
Process Overview of SAP MDM Consolidating Master Data - Example

SAP NetWeaver
? =
BI
Brown & Partner Inc. 248 Meadow Lane Drive Sacramento Ca. 95816 Brownes & Partner Brown Inc.
MEADOW Lane
San Diego
CA 93860
248 Meadow Lane Drive Sacramento Ca. 95819
Knowledge Mgmt.
CRM
Legacy
3rd Party
Legacy
Demo
Consolidating Master Data Challenges & Solutions

How can the data from different systems be compared at all? How can duplicates or identicals be determined? How can granted/may be/most probably not hits be distinguished? How can the system itself decide what cleansing cases to
automatically confirm as duplicates or identicals? present to the master data specialist to take a decision? automatically confirm as non duplicates or identicals?
MDM cleansing capabilities provide: Normalization capabilities to store unaligned data in a comparable format Object type related Matching algorithms to compare data sets Ranking on matching results to calculate matching scores Lower and Upper Score Thresholds to pre-decide on cleansing cases
SAP MDM 3.00 - Architecture
BI (BW 3.5) EP 6.0 Master Data Engine 3.00 Master Data Clients*
*i.e. SAP ERP, CRM, SRM
XI 3.00
technical routing structure mapping key-mapping cache (new 3.0)
master data administration process control (process chains) Inbound staging UIs logical routing
SAP Solution Manager 3.1
Content Integrator 3.00

matching engine key-mapping administration
Content Integrator - Process Overview

Master Data Client (MDC)
Business Partner 1 Business Partner 2 Business Partner 3 Business Partner 4 Product 1 Product 4 Product 3 Product 2 Business Partner 5
Master Data Server (MDS)

Content Integrator (CI)
XML-File RFC
Upload
Validation
Normalization Staging Area

Global Normalizer GetNormalizedAttributeKeys NormalizeObjects
Matching
GetMatchingCandidates CalculateScore
Cleansing
ID-Mapping
The explication you can find in the notes
Content Integrator Process Details

Staging Area RFC API XML-File
CI
1. Upload Validation
2. Normalization Global Normalizer writes
GetNormalizedAttributeKeys
reads
Object Store Table
NormalizeObjects 4. Cleansing 3. Matching
writes Index Table reads GetMatchingCandidates
ID-Mapping Table
API
writes
Mapping
reads
Score Table
writes
CalculateScore
Comparison
writes MDS
Generation of MDS-Objects
the Normalization is triggered again for the MDS Object (stops after writing into the Index Table)
Overview: Table View

Table Content during process
MDC-Objects with attributes
Object Store Table
Content end of process

MDS-Objects with attributes MDC-Objects without attributes (exist only to hold the references) normalized and indexed MDS-Data
normalized and indexed MDC-Data

Index Table
Score Table
MDC-Data that lies above the lower threshold with the related calculated score mapped MDC-Objects
for every new process the content is overwritten
ID-Mapping Table
MDS-Objects with the information, to what MDC-Objects they are mapped
Matching Strategy (Context) The Matching Strategy

contains settings to Normalizing Algorithm Matching Algorithm is defined dependent on the particular Business Objects the requirements of the customer is used within a Matching Context (e.g. Global Spend)
Normalization: The Process (Generally)

Staging Area XML-File RFC
CI
1. Upload
2. Normalization Global Normalizer writes
reads
Object Store Table
NormalizeObjects
writes Index Table
Global Normalizer
Makes the data comparable Perticular settings: Ingnore blanks, special characters or several punctuation marks Upper and lower case
Normalizing Algorithm
GetNormailzedAttributeKeys Finds the indices that are defined within the coding NormalizeObject Indicates the data and writes them into the index table
Settings for the Global Normalizer
Object Type
Upper or lower case Include or exclude the characters Insert the character you want to include or exclude. Enter the Source Path / Target Path Enter the Reference Path
Settings for the Global Normalizer

Upper or lower case This concerns the whole content of the array. Include or exclude the characters If you include a certain amount of characters, the other characters, that are not separately mentioned are automatically excluded. Insert the character you want to include or exclude Be careful with the character <blank>: It must not stand in the last position when listing several characters. The characters have also to be listed without any separator. Enter the Source Path/Target Path It corresponds to the array name of the source/target table. You can get it form the Development Guide at the Service Market Place. Enter the Reference Path You can get it from the object scheme displayed in the Development Guide at the Service Market Place.
Global Normalizer Object model diagram of the Content Integrator Business Partner
Reference Path
Source Path/Target Path
(source: Development Guide for Matching Strategies)
The indices can be defined within the java coding You can define as an index
one array several arrays a certain part of an array (e.g. the first 8 characters)
Due to the fact that you can manipulate directly in the coding, the specification of the indices is very flexible.
NormalizeObjects
The data are indicated and written into the index table, dependent from the previous coding settings:
Mandant Object ID consecutive numbering of data records having the same ObjectID Matched Entity ObjectID Column Key/Index Column Value
one object
NameNormalizer Customizing
At the SAP Enterprise Portal a Normalizing Algorithm Customizing in XMLFormat can be uploaded.
It may contain: The definition of NonNameTokens The substitution of several characters or a special character string The adding/truncation of several characters or a special character string
Settings for Normalizing Algorithm
Java Classname for Normalizing Algorithm
Customizing XML file according to the Normalizing Algorithm.
Example Normalization
Activity Example
"Mnchen Traffic Corp." 1 Remove fill signs like "# _ - / \ . ," "Mnchen Traffic Corp" 2 Convert to upper case "MNCHEN TRAFFIC CORP" 3 Normalize special characters according to predefined list (replace by AE) MUENCHEN TRAFFIC CORP 4 (Tokenize) Cut name into tokens using list of token separator: <blank>, "-" MUENCHEN, TRAFFIC, CORP 5 Check and mark token against predefined list of nonname tokens (like CORP, BANK). MUENCHEN, TRAFFIC, CORP 6 Check for minimum Length. MUENCHEN, TRAFFIC, CORP
Matching: The Process (Generally)

GetMatchingCandidates Reads the content of the Index Table CalculateScore Calculates the score between the indicated data records the score is calculated with the help of the settings in the xml-file dependent on the defined upper and lower threshold, the results are written into the Score Table objects with a score below the lower threshold are not written into the Score Table
4. Cleansing 3. Matching Index Table reads GetMatchingCandidates writes
Score Table
CalculateScore
Matching Algorithm
CalculateScore
Calculates the score between the indicated data records with the help of the settings in the XML-File
Matching Algorithm contains: the definition of the score for the particular attributes conditions for the matching check (e.g. check in a certain order with a special dependence)
The score values in a existing XML-File can be replaced without changing the Matching Algorithm.
Matching Strategy BP_300_Token_cust - Organizations
Source of Information Partner/companyName
Comment
Scoring +25 +10 +10 for first for each additional for each match of NameToken match of NameToken match of NonNameToken
Partner/SecondaryIDs/PartnerIdentification
Only for ABA BuPa available
+50 +20 10 +50 +20 10 -30 +15 +5 +10 +15 +0 +5 +5 +5
if secID and secIDType match else +0 for each additional match of secID / secIDType pairs if secID does not, but secIDType matches if secID and taxType match else +0 for each additional match of secID / taxType pairs if secID no match, but taxType matches if no match, if match, if match, if match, for first for each additional for each for each if sum 90 if sum 90 if sum > 190 else +0 else +0 else +0 else +0 match of NameToken else match of NameToken match of HouseNumber match of NonNameToken
Partner/SecondaryIDs/TaxNumber
Partner/address/countryCode Partner/address/postalCode Partner/address/region Partner/address/city Partner/address/street Partner/address/houseNumber
Score
Sum up scoring points
sum (90 + (sum90)/10) 100
Matching Algorithm (Example)

Threshold
Original Data Name Street City Postal Code DUNS Normalized Date nameTok en nameTok en non-nameTok en street street number city Postal Code DUNS Matching Result Calculated Score Final Score Result
Reference Object Mnchen Trafic Corporation Brezelstrasse 7 Mnchen 80331 4711 Normalized Values MUENCHEN TRAFIC CORP Brezelstrasse 7 MUENCHEN 80331 4711
Matching Canditate 1 Mnchen Rck AG Brezelstrasse 17 Mnchen 61525 4512
Matching Canditate 2 Mnchener Brtchen Corp. Brezelstrasse 17 Mnchen 61525 -
Matching Canditate 3 Mnchen Trafic Corporation Brezelgasse 17 Mnchen 80331 4711
Normalized Values Score Normalized Values Score Normalized Values Score MUENCHEN 25 MUENCHEN 25 MUENCHEN 25 RUECK 0 BROETCHEN 0 TRAFIC 10 AG 0 CORP 10 CORP 10 Brezelstrasse 15 Brezelstrasse 15 Brezelgasse 0 17 0 17 0 17 0 MUENCHEN 10 MUENCHEN 10 MUENCHEN 10 61525 0 61525 0 80331 15 4512 -10 0 4711 50
40 40 No Match Cleansing Case
60 60 Automatic Match
120 91
Settings for Matching Algorithm
Java Classname for the Matching Algorithm
Customizing XML file according to the Matching Algorithm
Overview Architecture Normalization Matching Implementation Considerations
Find Ways to identify Duplicates/Identicals

How to find records representing the same business object?
1. Using the name 2. Using identifying characteristics 3. Using other attributes Name1, Name2, .. D&B no., tax no., manufacturer part no. Street, zip code, place
4. Built decision matrix - verify with business using real life examples Name Street House No. Zip Code City Account Group Duplicate? No No? Yes? Yes
Know about the Quality of your Data

Examples of questions about the quality of master data
Which identifying characteristics are available in your systems and are correctly maintained there? Are these characteristics always available? What are the insignificant parts of a name for which a matching strategy is not required? Which data might contain typing errors? How big is the volume of data in the different systems? What level of similarity is required for a valid duplicate proposal? Is there a level of similarity above which a duplicate is certain to be found? Are the existing matching strategies sufficient to meet the requirements?
Be aware of exceptional Cases

Reality Check: Example - Wal-Mart as a customer
If the matching strategy is based on the name, there will be a number of superfluous duplicate proposals and therefore a large amount of effort will be required for the manual decision (clearing). This will also have a negative effect on the system performance.
Possible Solution Adjust limit values for generating a duplicate proposal Make Wal-Mart an insignificant name part Adjust the matching strategy: Are there other identifying attributes or keys that can be taken into account?
Use all Configuration Possibilities (before Developing)

Matching results can be influenced on 3 levels
1. Portal based configuration
build your Matching Context choose from the out-of-the-box Normalization/Matching Algorithms define your thresholds (lower/upper) configure the Global Normalizer
2.
XML-file based configuration

normalization (e.g. define additional non name tokens) alter the scores for the particular attributes
3.
Adapt a standard algorithms, develop own algorithms

develop your own JAVA coding for normalization and/or matching define your own customizing objects for your new algorithms follow the Matching Strategy Development Guide
Developing an new Matching Strategy

Steps of the development process
1. Prerequisites and Configuration of the Development Environment
Install and configure the required software SAP MDM 3.0 SP02 or higher NetWeaver Developer Studio 2.0 or higher
2.
Building your application logic

Define your project using NetWeaver Developer Studio Import the matching API and perform recommended settings for your project Design your matching and normalization strategy and implement the logic in Java classes using the interfaces provided by the Matching API (consider Matching Strategy Development Guide)
3.
Customizing and Deploying of a Matching Strategy

Define new customizing object types for matching and normalization algorithm and upload them on the SAP Enterprise Portal. Customization files are maintained in XML format.
Fine-Tune the Normalization/Matching Algortihms

Test different Matching Strategies
In MDM 3.00 multiple Matching and Normalization Algorithms can be applied on objects within one object type. The data is loaded once, but can be used for multiple normalization and matching processes according to different matching contexts. Generated ID-Mappings will be saved separately marked with the relevant Matching Context. One Matching Context can be marked for productive usage. The whole process can be reset all relevant table content will be deleted.
Summary Content Integrator (CI) is a component of SAP Master Data Management (MDM) to consolidate business data in a heterogeneous system landscape. The matching process is the core process to identify identical or similar business data objects. Matching context provides the logic and conditions for matching processes. With MDM some Matching Strategies are delivered. Matching Strategies can be configured and/or developed.
Further Information
Public Web:
www.sap.com SAP Developer Network: www.sdn.sap.com Master Data Management SAP Customer Services Network: www.sap.com/services/
Related SAP Education Training Opportunities

http://www.sap.com/education/ http://service.sap.com/okp
Related MDM Documentation

http://service.sap.com/instguides Installation and Upgrade Guides SAP Components SAP MDM (Configuration Guides, Matching Strategy Development Guide, Description of standard Matching Strategies)
Related Workshops/Lectures at SAP TechEd 2004

MDM201 MDM202 MDM203 MDM252 MDM253 How to Use Key-Mapping in SAP MDM for Reporting Configuration of Matching-Strategies for SAP MDM Customizing ID Mapping in SAP Solution Manager Connecting a New Client-System to SAP MDM SAP MDM: How to Model a New MDM Object Type Lecture Lecture Lecture Hands-on Hands-on
SAP Developer Network

Look for SAP TechEd 04 presentations and videos on the SAP Developer Network. Coming in December. http://www.sdn.sap.com/
Questions?
Q&A
Feedback
Please complete your session evaluation. Be courteous deposit your trash, and do not take the handouts for the following session.
Thank You !
Overview Architecture Normalization Matching Algorithm Implementation Considerations Appendix
Matching Strategies
Matching Algorithm (com.sap.ci.strategies)
Mat_300_GTIN_A Mat_300_GTIN_SPN_MPN_A
Matching Strategies CI 3.0

Mat_300_GTIN Mat_300_GTIN_SPN_MPN
Normalization Algorithm (com.sap.ci.strategies)

Mat_300_GTIN_N Mat_300_GTIN_SPN_MPN_N Mat_300_ShorttextCategory_cust_N Org_300_Token_cust_N BP_300_Token_cust_N TA_300_ManufacturerInfo_N MDMPartnerStrategyByTokens DummyMaterialNormalizationAlgo MaterialNormalizationByGTIN_SPN_MPN TechnicalAssetNormalization
Mat_300_ShorttextCategory_cust Mat_300_ShorttextCategory_cust_A Org_300_Token_cust BP_300_Token_cust TA_300_ManufacturerInfo Organization_By_Token Material_by_GTIN Material_by_GTIN_MPN_SPN TechAsset_by_ManuInfo

Org_300_Token_cust_A BP_300_Token_cust_A TA_300_ManufacturerInfo_A MDMPartnerStrategyByTokens MDMProductStrategyByGTIN MDMProductStrategyByGTIN_SPN_MPN MDMTechnicalAssetStrategyByManufacturerInfo
Please note that this document is subject to change and may be changed by SAP at any time without notice. The document is not intended to be binding upon SAP to any particular course of business, product strategy and/or development.
Cleansing: How to Merge Manually
Enter one of the data cleansing case numbers
List of Cleansing Cases: The mapped data records are displayed
To compare the two data, mark them and click the button <Compare Objects>
To merge the two objects: Set one of the data <Duplicate>, the other <Target> Set the status from <New> to <Released> Click the button <Save>
Additionally you have to enter the other cleansing case number.
Copyright 2004 SAP AG. All Rights Reserved

No part of this publication may be reproduced or transmitted in any form or for any purpose without the express permission of SAP AG. The information contained herein may be changed without prior notice. Some software products marketed by SAP AG and its distributors contain proprietary software components of other software vendors. Microsoft, Windows, Outlook, and PowerPoint are registered trademarks of Microsoft Corporation. IBM, DB2, DB2 Universal Database, OS/2, Parallel Sysplex, MVS/ESA, AIX, S/390, AS/400, OS/390, OS/400, iSeries, pSeries, xSeries, zSeries, z/OS, AFP, Intelligent Miner, WebSphere, Netfinity, Tivoli, and Informix are trademarks or registered trademarks of IBM Corporation in the United States and/or other countries. Oracle is a registered trademark of Oracle Corporation. UNIX, X/Open, OSF/1, and Motif are registered trademarks of the Open Group. Citrix, ICA, Program Neighborhood, MetaFrame, WinFrame, VideoFrame, and MultiWin are trademarks or registered trademarks of Citrix Systems, Inc. HTML, XML, XHTML and W3C are trademarks or registered trademarks of W3C, World Wide Web Consortium, Massachusetts Institute of Technology. Java is a registered trademark of Sun Microsystems, Inc. JavaScript is a registered trademark of Sun Microsystems, Inc., used under license for technology invented and implemented by Netscape. MaxDB is a trademark of MySQL AB, Sweden. SAP, R/3, mySAP, mySAP.com, xApps, xApp, SAP NetWeaver and other SAP products and services mentioned herein as well as their respective logos are trademarks or registered trademarks of SAP AG in Germany and in several other countries all over the world. All other product and service names mentioned are the trademarks of their respective companies. Data contained in this document serves informational purposes only. National product specifications may vary. These materials are subject to change without notice. These materials are provided by SAP AG and its affiliated companies ("SAP Group") for informational purposes only, without representation or warranty of any kind, and SAP Group shall not be liable for errors or omissions with respect to the materials. The only warranties for SAP Group products and services are those that are set forth in the express warranty statements accompanying such products and services, if any. Nothing herein should be construed as constituting an additional warranty.

Session ID: MDM202 Configuration of Matching-Strategies For SAP Master Data Management

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Session ID: MDM202 Configuration of Matching-Strategies For SAP Master Data Management

Hochgeladen von

Copyright:

Verfügbare Formate

Session ID: MDM202 Configuration of Matching-Strategies for SAP Master Data Management

SAP AG 2004, SAP TechEd / Session MDM202 / 2

As a result of this workshop, you will be able to:

SAP AG 2004, SAP TechEd / Session MDM202 / 3

Overview Architecture Normalization Matching Algorithm Implementation Considerations

MDM A Key Capability of SAP NetWeaver 04

SAP NetWeaver MDM

augment, and consolidate

Multi channel access Composite Application Framework Portal Collaboration

Bus. Intelligence Master Data Mgmt

Business Process Mgmt

execution of business processes

SAP AG 2004, SAP TechEd / Session MDM202 / 5

SAP NetWeaver: Delivering on the promise of ESA

INFORMATION INTEGRATION Knowledge Mgmt Bus. Intelligence

APPLICATION PLATFORM J2EE ABAP

SAP AG 2004, SAP TechEd / Session MDM202 / 6

Process Overview of SAP MDM Loading Master Data

ERP Application Platform

SAP AG 2004, SAP TechEd / Session MDM202 / 7

Process Overview of SAP MDM Consolidating Master Data

ERP Application Platform

Consolidating Master Data from Staging (PIC)

Valid for both process variants

SAP AG 2004, SAP TechEd / Session MDM202 / 8

Process Overview of SAP MDM Consolidating Master Data - Example

248 Meadow Lane Drive Sacramento Ca. 95819

ERP Application Platform

Consolidating Master Data Challenges & Solutions

SAP AG 2004, SAP TechEd / Session MDM202 / 10

Overview Architecture Normalization Matching Algorithm Implementation Considerations

SAP MDM 3.00 - Architecture

SAP Solution Manager 3.1

Content Integrator 3.00

SAP AG 2004, SAP TechEd / Session MDM202 / 12

Content Integrator - Process Overview

Master Data Server (MDS)

Normalization Staging Area

The explication you can find in the notes

SAP AG 2004, SAP TechEd / Session MDM202 / 13

Content Integrator Process Details

2. Normalization Global Normalizer writes

Object Store Table

NormalizeObjects 4. Cleansing 3. Matching

writes Index Table reads GetMatchingCandidates

SAP AG 2004, SAP TechEd / Session MDM202 / 14

Overview: Table View

Content end of process

normalized and indexed MDC-Data

for every new process the content is overwritten

MDS-Objects with the information, to what MDC-Objects they are mapped

SAP AG 2004, SAP TechEd / Session MDM202 / 15

Matching Strategy (Context) The Matching Strategy

SAP AG 2004, SAP TechEd / Session MDM202 / 16

Overview Architecture Normalization Matching Algorithm Implementation Considerations

Normalization: The Process (Generally)

2. Normalization Global Normalizer writes

Object Store Table

writes Index Table

Settings for the Global Normalizer

Settings for the Global Normalizer

SAP AG 2004, SAP TechEd / Session MDM202 / 20

Source Path/Target Path