Launching Semantic Web Launching Semantic Web Technology in Enterprise IT

handclockΑσφάλεια

5 Νοε 2013 (πριν από 3 χρόνια και 10 μήνες)

115 εμφανίσεις

we don’t make it
we make ITbetter
LaunchingSemanticWeb
Launching

Semantic

Web

Technology in Enterprise IT
Lessons Learned from Leading/Managing Semantic Web Projects
Jenna Zhou, Chary TamirisaJune, 2011
A little bit about us…
•Jenna Zhou
–Currently, enterprise architecture consultant
10yearexperiencewithenterpriseinformationarchitecture

10

year

experience

with

enterprise

information

architecture
–CDMP and CBIP (both mastery level)
–IASA certified architect
–M.S. in Biochemical Engineering, Chemical Engineering, Computer
Science and MB
A
•Chary Tamirisa
Currentlyenterprisearchitectureseniorconsultant

Currently
,
enterprise

architecture
,
senior

consultant
–Over 28 years of experience in Software Engineering
–Expert in Enterprise Application Integration, Distributed Computing,
Business Intelligence, Multithreading
IASAtifidhitt

IASA
cer
tifi
e
d
arc
hit
ec
t
–M.S. in Bioinformatics, Computer Science
we don’t make it
we make ITbetter
2
What about you?
•How much experience do you have with semantic web technology?
Aidilfthilil?

A
re you cons
id
er
i
ng yourse
lf
more
t
ec
h
n
i
ca
l
or more manager
i
a
l?
we don’t make it
we make ITbetter
3
Agenda
•Potential Usages of Semantic Web Technology

PotentialBarriersforAdoption
Potential

Barriers

for

Adoption
•What’s Unique in Enterprise IT
•Where to Start
•Example: Enterprise Integration
•How to Manage Semantic Web Projects

LessonsLearned

Lessons

Learned
•Q & A
we don’t make it
we make ITbetter
4
Potential Usages
Semantic
Annotation
Semantic
Linked Data
Annotation
Semantic

Master Data
Management
Semantic
Integration
Ontology
Driven
Architecture
Bring the
tit
Semantic
seman
ti
cs
t
o
the data!
Search
Engine

we don’t make it
we make ITbetter
5
Potential Barriers
New Technology
Reality
Unknown
Fear of the unknown
risk averse
Uncomfortablewith
Requires
change
Uncomfortable

with

change
unable to change
Requires
learn
No time to learn
Resource limitation
Lack of relevant
references
Rely on references and
experiences
we don’t make it
we make ITbetter
6
What is unique about Enterprise IT?
•The landscape consists of well tested
productionqualitysolutions
production

quality

solutions
•Architecture review board to govern
deployedtechnologies
deployed

technologies
•Emphasis on reliability, not innovation
we don’t make it
we make ITbetter
7
Where to Start?
Innovation
Risk
Reward
we don’t make it
we make ITbetter
8
8
we don’t make it
we make ITbetter
What
differentiates
Whith
What

differentiates
semantic web
technolo
gy
?
Wh
ere
i
s
th
e
business value
?
gy
we don’t make it
we make ITbetter
9
Value Propositions to Start With
Level 3
Machine
Machine

Readable /
Process-able
Semantics
Pfit
Gth
Level 2
Relationships &
Contexts
P
ro
fit
G
row
th
Level1
Efficiency
Risk
management
Level

1
Terms and Definitions
we don’t make it
we make ITbetter
10
What differentiates semantic web technology?
•Resources are uniquely identified by URIs

Propertiesarefirstlevelclasses
too!

Properties

are

first

level

classes

too!
•Simple and flexible triple structure

Descriptivelogic

inference
Descriptive

logic

inference
•Query is based on pattern match

Openworldassumption
Open

world

assumption
•Machine readable and actionable ontology
we don’t make it
we make ITbetter
11
Example: Enterprise Integration
ETL Approach:
DataSource1
SA
Every connection is a data mapping
Data

Source

1
Data Source 2
DataSource3
Target
S
ystem
A
Target System B
Data Source …
Data

Source

3
Target System …
SOAAh
SOA

A
pproac
h
::
Data Source 1
DataSource2
Canonical Model A
Service I
DataSource…
Data

Source

2
Data Source 3
Canonical Model B
Canonical Model …
Service II
Service …
we don’t make it
we make ITbetter
12
Data

Source


Example: Enterprise Integration (Contd.)
With traditional approaches
•Every connection is a data mapping
•Most data mapping is manual work
•Most data mapping is kept in spreadsheet
•Sometimes, data mapping is not achievable
•Minimal reuse
•Difficult for collaborative develo
p
ment
p
•Low productivity
•May lose semantics
E

E
rror prone
•Impossible for effective governance
we don’t make it
we make ITbetter
13
Example: Enterprise Integration (Contd.)
Uitibthl
hasWorkedOn
U
s
i
ng seman
ti
c we
b

t
ec
h
no
l
ogy
hasWorkedOn
isProjectLeade
r
isTestingAssociate
isQualityAssociate
isTeamLeade
r
isTechnicalLeade
r
isLeadDeveloper
isLeadSoftwareEngineer
we don’t make it
we make ITbetter
14
Example: Enterprise Integration (Contd.)
With semantic web technology
•Mapping to common ontology
•Data will not be forced into a fixed structure
•Consistency can be built in the ontology
•Class and property can be explicitly defined
•Mapping can be exported in reusable format
•Increase reuse
•Enable collaborative develo
p
ment
p
•Increase productivity
•Semantics will be kept
Rdt

R
e
d
uce error ra
t
e
•Facilitate effective governance
we don’t make it
we make ITbetter
15
Exercise: How would we communicate value?
Level 3
Mhi
M
ac
hi
ne
Readable
Semantics
Pfit
Gth
Level 2
Relationships &
Contexts
P
ro
fit
G
row
th
Level1
Efficiency
Risk
management
Level

1
Terms and Definitions
we don’t make it
we make ITbetter
16
Managing Semantic Web Projects
•Socialize the semantic web technology

Picktherightscope

Pick

the

right

scope
•Form the team

Managetheexpectations
Manage

the

expectations
we don’t make it
we make ITbetter
17
Lessons Learned
•Socializing the concepts

Workwithrealworldproblems

Work

with

real

world

problems
•Manage the appropriate expectations

Ensuretherightscopeandvalueproposition
Ensure

the

right

scope

and

value

proposition
•Start with good tools

Showtheroadmap
Show

the

road

map
•Be persistent
we don’t make it
we make ITbetter
18
Questions?
Questions?
we don’t make it
we make ITbetter
19
we don’t make it
we make ITbetter
THANK YOU!
Contact Information
Jenna_Zhou@dell.com
Chary_Tamirisa@dell.com
21
we don’t make it
we make ITbetter