D:
Key type, Key width and key type
Which one does over match if we are matching 1 lack records,
What are new features added compare to previous version.
What is tokenization. How it starts before match and merge and on what bases its matches.
How token ID’s generates.
Load by rowed object concept. And if I am getting src rowd same and pkey src object is
different . This condition how to overwrite existing record with this.
How do you identify date difference using cleanse function. Sysdate and date coming from
landing.
How do you configure message quee when I want to publish only USA data got updated.
This can be done by creating queries and packages for only US country and configured same
package in MQ.
Relationships in ID. What is part and all.
Configure lookup in IDD. Lookup by forein key value .. this can be done by creating relations in
MDM HUB. Go to BO and create relationship for particular column.
I want to search only USA and CANADA data in IDD. Subject area include both filter.
How many levels Childs configured in IDD.
What is time line concept.
What is smart search and 360 degree view.
When created HM, which are tables gets create.
If we convert base object to entity in HM. What are all columns get create in entity table.
Base object class code
Don’t load data when phone number has same phone number exists in BO.
What is segment matching.
Go through all cleanse functions. Including reject and readdatabse.
If record not participating in match and merge because it is already consolidated record, so
when new record comes to match .
Then how new record matches with consolidated record.
Survivor is consolidated record and victim is new record. When match and merge runs first insert
tokens into STRP table. Then once match and merge done. It will delete victim tokens from STRP table
and also deletes from DRTY table, but survivor record(consolidation record) tokes will be available in
both tables.
So when next run it will generate tokes for consolidated record and for new record which is going to
merge. But will not participate in match and merge only new record matches with this record.
[Link]
For exact match new record matches with all records in the table(Data match). For Fuzzy match one
tokes is matches with all rest of tokes to identify the target rowed object.
Flow chart:
Tokenization Match process Consolidation Process Merge.
How to optimize match and merge.
Fuzzy match algorithm CMOS.
Read properly SAM.
What is over match and under match . How to configure this.
Match Level: Indicates the accuracy level of the fuzzy match. It has the following options:
Typical: Suitable and applicable for most of the matches
Conservative: Produces fewer matches relative to Typical level (this is called
“Undermatching”).
Loose: Produces more matches relative to Typical level (this is called as “Overmatching”).
If two records are merged . Then lowest rowed object will survive.
HM:
Create queries and packages for the RBO tables.
Note down columns which are created for the tables. Like Base object class code
Add entity by converting BO to entity. Then note down what are columns inserted in the
entity table. BO_CLASS_CD inserted to BO table.
Entity created data will get stored in RBO tables.
Create relationship between BO table and RBO tables. This can be done in staging table by
creating lookup value set up with RBO table with the column BO_CLASS_CD.
Create new entity for reporting. If you create new entity then it will create one BO with
same name and with BO_CLASS_CD column. Then add all required columns for reporting.
Create entity [Link] entytype created then insert data into all RBO tables.
Add hierarchy in hierarchy tab.
Then create relationship between entity and Hierarchy. By using options like create new and
convert existing BO relationship.
Create reporting tables.
Create foreign key relationship between BO in the Hierarchy tab.
Add relationship type .
Create packages for all relationship tables. Go to profile and create packages.
Create profile and select pakcge and select columns to display and assign priority.
Tool Tip in MDM:
If you want to alert user with message whenever empty name.
Create cleanse function to validate like is empty cleanse function.
Then configure same celanse function in IDD for particular column.
Then add string in [Link] bundle file. Message which you want to see.
Then deploy.
1. Mdm implementation styles
2. Mdm architecture
3. Mdm modules
4. Cleanse list d cleanse function diff
5. How to delete records from bo when data is not in landing but exist in bo. First load loaded but in ne
6. How to generate sequence number in bo column using function
7. What is fuzzy key.. what is the use
8. What is tokenization
9. How data gets loaded to all supporting table like prl,reject,raw
With example
10. Validation rules to downgrade trust.. restricted to downgrade 50 .. should not go beyond [Link]
11. Downgrade trust by 40 when fn null and 20 when last name null. If both are satisfied to one record
12. After defining relationship, what will you do.
13. What r the primary conditions to define relationship.
14. What is match d merge process.
15. Which engine using for 10.1
16. Hub installation process
17. How many sub levels can define in did
18. How many database hub support like Oracle d etc
19. Message que
20. Hierarchy
21. Miscellanea functions list
22. All other inbuilt functions list
23. Don't want to populate the records in staging with the fn is puru. How to handle this in land to xref j
24. How many types of rejection.
25. What r rejection from stage to load
What is HDD
How do u define lookup in staging
Explain about module
What is many to many relationship
Which are all analytics functions
How do u stop the job (clear c_repos_job_control table)
How do identify changes in mdm while importing d exporting
What is raw and prl table
How delta detection will work
Java
How to define many to many relationship
Explain data model
1. What is staging table
2. What is golden record.
3. What is fuzzy match
4. Is it possible to define exact match column to fuzzy Mach
5. Is it possible to map one staging table to multiple landing tables
6. How do u define job group
7. Which tables like lookup will load first.. how do u define in job group
[Link] it possible to load data to parent if data is not available in child.
9. Look relationship
10. If person told to load data. WhT r the requirements u need.
11. Is it possible to define percentage match like 60 70
12. How do you handles if phone number is 9 digits.
13. Is it possible to load landing data from hub.
14. If we know highest trust system, then why do we need other sources data.
15. 10 hub version called as hot fix 1
1. Segment match
2. Hmrg table . Hw to identify unmerge in hmrg table.
3. What is null match enable
4. Dirty Ind column in which tables
5. In which process validation rule trigger
6. What is difference between xref d Bo
7. What r difference between 9 d 10 version
8.
1. During load job hw to identify new d old records
[Link] has loaded but child has not loaded then how to create data relationship
[Link] records are merged , then which one will take priority(older rowid)
[Link] we r defining path
[Link] r issues faced during idl
[Link] job failed what will do
[Link] new column added , then what r all the things ..Will take care
[Link] frequently running incremental
[Link] many records will load per day
[Link] records inserted through idd , then how cleanse d lookup will handle
[Link] is dirty indicator
[Link] new d old records both consider in march d merge( only new record)
1. How to reject records using cleanse
2. Is tr possible to prepare query's on xref
3. Match by primary key in match rule setup
4. How to setup IDD .. SET UP BY MATCH PATH
5. How to make restrictions to Text feild in IDd [Link] entering text for numeric feild. Done by usin
[Link] records comes for update, the change is for match defined column. Then what will happen
7. All the token I'd will be same for same records in tokenization
8. What about match key
9. Explain fuzzy match process
[Link] key distribution. Match key will be same for all the matched records
1. How to set up many to many relationship in MDM
2. How to skip tokenization
3. What is difference between segment and segment all.
4. How to validate data in mapping
Like email phone number
Link to video
[Link]
5. How set dynamic trust setup
6. What is null matches non null
7. What are match rule type's.. narrow, typically, exhaustive, extreme
8. Match level typical, conservative and loose
9. HM
10. How to configure idd
11. What are sub group are types like part
12. What is 360 digree view
13. How to create role based login to idd
14. Match algorithm
Key type, Key width and key type
Which one does over match if we are matching 1 lack records,
What are new features added compare to previous version.
What is tokenization. How it starts before match and merge and on what bases its matches.
How token ID’s generates.
Load by rowed object concept. And if I am getting src rowd same and pkey src object is different . This condition h
How do you identify date difference using cleanse function. Sysdate and date coming from landing.
How do you configure message quee when I want to publish only USA data got updated.
Relationships in ID. What is part and all.
Configure lookup in IDD.
I want to search only USA and CANADA data in IDD. Subject area include both filter.
How many levels Childs configured in IDD.
What is time line concept.
What is smart search and 360 degree view.
When created HM, which are tables gets create.
If we convert base object to entity in HM. What are all columns get create in entity table.
Don’t load data when phone number has same phone number exists in BO.
What is segment matching.
Go through all cleanse functions. Including reject and readdatabse.
If record not participating in match and merge because it is already consolidated record, so when record comes t
Then how new record matches with consolidated record.
How to optimize match and merge.
Fuzzy match algorithm CMOS.
Read properly SAM.
ReplyForward
Puru G
Global Data and Analytics, Data Science / JJT TS
#152, ITPL Main Road, Bengaluru, KA 560066 India
Mobile: +91 9742499879