Más contenido relacionado La actualidad más candente (20) Similar a Knowledge Graphs, Ontologies, and AI Applications (20) Más de Earley Information Science (20) Knowledge Graphs, Ontologies, and AI Applications1. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved.
KNOWLEDGE GRAPHS,
ONTOLOGIES AND AI
KM World
Seth Earley
WWW.EARLEY.COM
@sethearley
seth@earley.com
www.linkedin.com/in/sethearley
2. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved.
Seth Earley - Biography
CEO and Founder
Earley Information
Science
@sethearley
seth@earley.com
www.linkedin.com/in/sethearley
Over 20 years experience
Current work
Co-author
Editor
Member
Former Co-Chair
Founder
Former adjunct professor
Speaker
AIIM Master Trainer
Course Developer & Master Instructor
Data science and technology, content and knowledge
management systems, background in sciences (chemistry)
Enterprise IA and Semantic Search
Information Organization and Access
Industry conferences on knowledge and information management
Northeastern University
Boston Knowledge Management Forum
Academy of Motion Picture Arts and Sciences, Science and
Technology Council Metadata Project Committee
Editorial Journal of Applied Marketing Analytics
Data Analytics Department IEEE IT Professional Magazine
Practical Knowledge Management from IBM Press
Cognitive computing, knowledge and data management systems,
taxonomy, ontology and metadata governance strategies
2
3. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved.
The AI Powered Enterprise
3
Available at
https://www.amazon.com/AI-Powered-
Enterprise-Ontologies-Business-
Profitable/dp/1928055508/
“A great resource to
separate the hype from the
reality and a practical guide
to achieve real business
outcomes using AI
technology.”
—Peter N Johnson, MetLife
Fellow, SVP, MetLife
“I do not know of any books
that have such useful and
detailed advice on the
relationship between data and
successful conversational AI
systems.”
—Tom Davenport, President’s
Distinguished Professor at Babson
College, Research Fellow at MIT
Initiative on the Digital Economy,
and author of Only Humans Need
Apply and The AI Advantage
“Read this book to learn how
leaders and companies are
using AI with structured data
to transform business. Insight
from real world examples,
combined with a proven
methodology, will arm the
reader with the knowledge
and confidence necessary to
drive AI in any organization”.
– Barry Coflan, SVP & Chief
Technology Officer, Schneider
Electric – Digital Energy
4. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved.
Agenda
5
1. What are knowledge graphs? Why the hype?
Is it justified?
2. How are knowledge graphs leveraged in the
enterprise?
3. How can knowledge graphs power AI
applications?
www.earley.com @sethearley
5. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved.
Market Expectations and Communications*
6
“Knowledge graphs (KGs) solve well-known data and content management problems.
KGs are the ultimate linking engine for enterprise data management.
KGs automatically generate unified views of heterogeneous and initially unconnected
data sources, such as Customer 360.
KGs provide reusable data sets to be used in analytics platforms or to train machine
learning algorithms.
KGs help with the dismantling of data silos. A semantic data fabric is the basis for
more detailed analyses”
www.earley.com @sethearley
* Hype – the motherhood and apple pie of what
everyone wants from technology
Source: The Knowledge Graph Cookbook https://www.poolparty.biz/the-knowledge-graph-cookbook/
6. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved.
Confusion
7
Enterprise graph
Entity relationships
Property graphs
Labeled properties
Labeled property graph
Nodes, Entities, Edges
Attributes on edges
Schema federation
Constraint management
Semantic graphs
Inference using RDF, RDF*,
OWL, SPARQL
If you want executive funding and support, don’t:
Or use language like this:
Show diagrams like this:
Instead, demonstrate capabilities and show
measurable business outcomes
7. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved.
Graph, Graph Data, Knowledge Graph
Graph – mathematical representation of objects (called a node) and
relationships (called an edge)
Graph Database – focus on the relationships between data points
rather than the data itself
Knowledge Graph – representation of unstructured content
categorized across multiple metadata elements.
8
Knowledge Graphs (and more broadly graph data) allow for contextual
navigation across an unstructured repository of artefacts and the linkage
of disparate data sources based on common elements or attributes
8. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved.
Enterprise Information
Challenges
THE ROLE OF GRAPH DATA AND
KNOWLEDGE GRAPHS
9
www.earley.com @sethearley
9. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved.
Too many tables and
attributes
Impossible to
understand naming
Complex
Relationships
Data is application
centric
Data experts
unavailable
Documentation
non-existent
Master databases
are off limits
Data quality
unknown
The enterprise architect’s dilemma
Source: Juan Sequeda, data.world
10
10. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved.
The Integration, Navigation, Retrieval Challenge
11
Order
Management
ERP CRM Support eCommerce
Data Data Data Data Data
Customer
Content
Contract
Customer
Content
Contacts
Account Customer
Personas
Product
Product Contact Info
Customer
Orders
Product
Content
Customer
Prospect
Content
Operations
Data
BOM
Content
PLM
Content
Fragmented systems
Disconnected processes
Unclear ownership
Unmanaged lifecycle
11. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved.
The user’s dilemma – “where do I find…”
SAP BPN
DAM
ISSUES
Diagnostics
FIELD
NOTIFICATION
COLLABORATION
CRM
MARKETING
ASSETS
12
Oracle
EXPERTISE
LOCATION
TECH PUBS
LIBRARY
SHAREPOINT
LIBRARIES
12. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved.
What is every project’s answer to
application proliferation?
Another application!
“if we just had one place for everyone to go…”
“we can migrate to a central location…”
“we need migrate all of our content and data to a
repository where all of our people can find their stuff
…”
13
13. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved.
Information Sources and Retrieval are Varied
ONTOLOGY BASED INTEGRATION FRAMEWORK
SOURCES
RETRIEVAL
BI Integration
Auto categorization/
Clustering
Entity
Extraction
Faceted
Search
Semantic
Search
Business Intelligence
Customer Relationship Mgt
Document repositories
Custom databases and applications
Intranets/web pages
Product Lifecycle Management Digital Asset Management
Data Warehouses
Messaging
ERP Systems
Knowledge Graph Navigation
Collaboration Spaces
14
DEVICES
KNOWLEDGE GRAPH
NAVIGATION BASED ON
UNIFIED ONTOLOGY
FRAMEWORK
14. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved.
BI, KM and KG
Business
Intelligence
Knowledge
Management
Knowledge Graph
Nature of
information
Structured Unstructured Unstructured and
Structured
Mechanism Retrieving Data from
disparate sources
Retrieving Content from
disparate sources
Retrieving Data and
Content from disparate
sources
Insights What happened? Why did it happen? What happened and why
did it happen?
Enhancement Database joins Content enrichment Joins on unstructured
content sources
Strengths Ability to deal with large
volumes of data, many
tools already in place
Ability to auto-categorize
content and provide
associative relationships,
ability to leverage search
platform
Ability to provide flexibility
of ad hoc queries across
systems using attributes
from structured and
unstructured sources
Weaknesses or
drawbacks
Inability to natively
connect to unstructured
data and lack of content
enrichment mechanisms
Inability to perform
database joins and process
large amounts of data
Addition of new tools to
enterprise
15
15. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved.
Graph Data vs Knowledge Graph
16
Graph Data: focus on relationships between elements
Knowledge Graphs: representations of unstructured information categorized and classified across
multiple metadata elements.
For example, a movie (described in data terms generically as a “production”) has the following metadata
attributes (“is-ness” and “about-ness”)
Title
Producers
Directors
Stars
Release
Synopsis
MPA
Type
Cast
Writers
Critic
Audience
Awards
Genre
Is-ness = “Production”
A Production has the following
descriptors: title, production type,
producers, MPA rating, directors,
synopsis, stars, genre, etc.
About-ness = attributes of
“Production”
Production
16. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved.
IMDB Graph Data
17
One object’s “is-ness” can be another object’s “about-ness”
“Is-ness”
If a “movie” has an “award”,
the award is an attribute of
that movie
17. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved.
Relationships Between Is-ness and About-ness
18
Feature film
Director of
Writer on
Actor in
Crew of
Producer of
Person
Title
Producers
Directors
Stars
Release
Title
Synopsis
MPA R
Type
Cast
Writers
Critic Score
Audience Sc
Awards
Genre
Production
Title
Surname
Role
Birthdate
Birthday
Name
Groupings
Awards
TV series
TV episode
TV movie
Video
Short film
TV mini-series
TV short
TV special
Video game
Etc.
Etc.
Birthplace
18. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved.
Relationships Between Is-ness and About-ness
19
Movie Award
Has director is the
about-ness of “movie”
Award is for a movie
(about-ness of award)
Director
Movie has an award
(about-ness of movie)
Directed a movie is the
about-ness of “director”
TV Show
Crew
Cast
Writers
Producers
Has producer, has cast, has
crew, has writers, etc. are all
about-ness of a “movie”
Award
Producer of
Cast of
Crew of
Writer of
about-ness of
“producer”, of “cast”,
of “crew”, or of “writer”
19. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved.
IMDB Graph Data
20
20. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved.
Linking Data Based on Attribute
21
21. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved.
Linking Data Based on Attribute
22
22. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved.
Linking Data Based on Attribute
30 pages!
23
23. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved.
Linking Data Based on Relationships and Attributes
Plus:
Sound department
Special effects
Visual effects
Stunts
Camera and electrical
Animation
Casting
Costume and Wardrobe
Editorial
24
24. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved.
Graph Data Advantages
• Ability to use simple data models to create complex reporting and look
ups leveraging data relationships
• Faster performance than database joins
• Ability to integrate disparate data sources
• Knowledge graphs are contextual in nature – by understanding
relationships, context is inferred
Knowledge graphs and graph data power AI and Machine Learning
systems by providing reference data and knowledge about conceptual
relationships between products, solutions, problems, tasks and
processes
25
25. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved.
Ontologies are the structural frameworks for organizing information and
are used in artificial intelligence, the Semantic Web, systems engineering,
software engineering, library science, and information architecture as a
form of knowledge representation.
The creation of domain ontologies is also fundamental to the definition and use
of an enterprise architecture framework.
Ontologies and Graph Data
26
www.earley.com
Source: https://datafloq.com/read/role-ontology-plays-big-data/749
Ontologies contain relationships that are expressed in knowledge graphs
Knowledge graphs include the actual data as represented in ontologies
26. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved.
Ontology is the knowledge
scaffolding of the enterprise
Graph data fills in that scaffolding
with the operational and
transactional data of the organization
27
27. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved.
Knowledge
graphs and
ontology are at
the core of a
unified
integration
framework
28
Integration
Framework
(Common/ mapped ontology
and metadata)
Predictive Analytics
Cognitive Technologies
Business Intelligence tools
Data Visualization Applications
STRUCTURED
DATA
UNSTRUCTURED
DATA
STRUCTURED
CONTENT
UNSTRUCTURED
CONTENT
SENSOR DATA
LOG FILES
CLICKSTREAMS
SOCIAL MEDIA
VOICE OF THE
CUSTOMER
ERP
DATA MARTS
CRM
DIGITAL MARKETING
PRODUCT DATA &
CONTENT
28. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved.
Ontology is
the
contextual
and semantic
framework
for the
enterprise
Knowledge Graphs
express and apply
enterprise data using the
ontology framework
29
The Bottom Line
29. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved.
Ontology Expressed as Graph Data Provides Consistent Architecture
30
COMMON ENTERPRISE ARCHITECTURE
Context Aware Information Architecture
Content Model Ontology Metadata
Structured
(Operational) Data
Unstructured
(Big) Data
Information Infrastructure
Marketing
Data
User
Data
Product
Data
Historical
Data
Operating
Content
Information Management Platforms
PIM DAM CMS ECM CRM ERP
Customer
Personalization
Content
Publishing
Site
Merchandizing
Product Info.
Management
Digital Commerce
Business
Intelligence
Knowledge
Management
Enterprise Search
Content
Management
Digital
Workplace
30. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved.
Caveats
• Knowledge graphs do not fix bad data
• Knowledge graphs require up front work establishing relationships
• Some graph relationships can be derived, but human judgment is still required
• Knowledge graphs provide context for cognitive and machine learning
applications
• Contextualization can then support recommendation systems, 360-degree view
of customers and data integration fabrics for unified information access
• Data and information governance, data quality and metrics driven decision
making frameworks are required to move the needle on enterprise initiatives –
using both conventional and emerging (AI powered) technologies
31
Knowledge graphs are the next iteration of data
integration but there is still no magic
31. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved.
The AI Powered Enterprise
32
For the next 10 KM World
Attendees to send me your
information (mailing
address) I will send a free
signed copy of the book.
“This book provides
prescriptive guidance in the
context of real business
case studies to drive
success instead of
disappointment.”
—Peter N Johnson, MetLife
Fellow, SVP, MetLife
“Artificial intelligence holds
the power to transform your
business, and your career,
but there will be plenty of
challenges along the way.
Earley demystifies the topic
and provides a practical
roadmap for applying smarter
processes and technologies
across the enterprise. Now is
the time to explore AI, and
this book is a great place to
start.”
– Paul Roetzer, Founder & CEO,
Marketing Artificial Intelligence
Institute and author of The
Marketing Performance Blueprint
(My ask in return is that you
write an Amazon review.)
Send a note to:
Seth@earley.com
and copy:
Carolyn.Southwick@earley.com
32. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved. 33
Publication Topic URL
HR Professional
Magazine
Book Look – AI Powered
Enterprise by Seth Earley
https://hrprofessionalsmagazine.com/2020/10/27/book-look-ai-
powered-enterprise-by-seth-earley/
Supply Chain Quarterly How AI can improve supply
chains
https://www.supplychainquarterly.com/articles/3978-how-ai-and-data-
science-can-improve-supply-chains
CEOWORLD The Secret To Making Digital
Transformations Work
https://ceoworld.biz/2020/10/06/the-secret-to-making-digitals-
transformations-work-its-the-data/
Applied Marketing
Analytics
Au Bon Pan CIO book
review
https://www.dropbox.com/s/whqq6lz251u75kv/Applied_Marketing_An
alytics_Review-The-AI-Powered-Enterprise-Seth-Earley.pdf?dl=0
Document Imaging
Report
Without IA, There is no AI https://www.dropbox.com/s/n1qkvld1a03qjla/Document-Imaging-
Report-Interview-2020-09-18.pdf?dl=0
TFIR Insights AI Has Failed To Deliver On
Its Promises
https://www.tfir.io/ai-has-failed-to-deliver-on-its-promises-seth-earley/
Destination CRM Required Reading: AI and
the Customer Experience
https://www.destinationcrm.com/Articles/CRM-
Insights/Insight/Required-Reading-AI-Has-Complete-Power-Over-the-
Customer-Experience-142584.aspx
Future of Field Service Is AI Delivering On Its
Promise?
https://www.futureoffieldservice.com/2020/08/10/is-ai-delivering-on-
its-promise/
Business Focused Articles About Ontology and AI
33. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved. 34
Publication Topic URL
HR Executive As remote work continues,
what is AI’s role?
https://hrexecutive.com/as-remote-work-continues-what-is-ais-
role/
ClickZ How your organization can
become AI powered
https://www.clickz.com/how-your-organization-can-become-ai-
powered/262306/
SHRM The Future of AI Powered HR https://blog.hrps.org/blogpost/The-Future-of-AI-Powered-HR
CEOWORLD Ontology: The Key to Unlocking
the Power of AI
https://ceoworld.biz/2020/07/06/ontology-the-key-to-unlocking-the-
power-of-ai/
The Enterprisers
Project
Artificial Intelligence: 8 habits of
successful teams
https://enterprisersproject.com/article/2020/6/artificial-intelligence-
ai-8-habits-successful-teams
HR.com Using Artificial Intelligence To
Improve Recruiting
https://www.hr.com/en/magazines/talent_acquisition/june_2020_ta
lent_acquisition/using-artificial-intelligence-to-improve-
recruitin_kbjh2zs2.html
DemandGen 5 Easy Tips For Implementing
AI Into Your Marketing Mix
https://www.demandgenreport.com/blog/a/5-easy-tips-for-
implementing-ai-into-your-marketing-mix
InBusiness Harness the Power of AI https://inbusinessphx.com/technology-innovation/harness-the-
power-of-ai#.XuoZi0VKiUl
Future of Field Service
podcast
Podcast: The AI-Powered
Enterprise
https://www.futureoffieldservice.com/2020/05/27/the-ai-powered-
enterprise/
Business Focused Articles About Ontology and AI
34. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved. 35
Publication Topic URL
TechTarget AI's impact on business https://searchenterpriseai.techtarget.com/feature/AIs-impact-on-
business-The-quest-to-make-money
Business Class
News
Podcast: Delivering on the
promise of AI
https://businessclassnews.com/books/delivering-on-the-promise-of-
artificial-intelligence/
HR Executive Exploring AI? https://hrexecutive.com/exploring-ai-plan-understand-and-control/
TechTarget 6 key benefits of AI for
business
https://searchenterpriseai.techtarget.com/feature/6-key-benefits-of-AI-
for-business
HR.com Building Effective Intranet
Systems
https://www.hr.com/en/magazines/all_articles/the-essential-role-of-hr-
in-building-effective-int_k9jihnnn.html
Harvard Business
Review
Is Your Data Infrastructure
Ready for AI?
https://hbr.org/2020/04/is-your-data-infrastructure-ready-for-
ai?ab=hero-main-text
TechTarget Importance of AI in the
business quest for data-driven
operations
https://searchenterpriseai.techtarget.com/feature/Importance-of-AI-in-
the-business-quest-for-data-driven-operations
TechTarget 5 AI Risks Businesses Must
Confront and How to Address
Them
https://searchenterpriseai.techtarget.com/feature/5-AI-risks-
businesses-must-confront-and-how-to-address-them
C-Suite Network Episode on Best Seller TV https://c-suitenetwork.com/tv/video/seth-earley-the-ai-powered-
enterprise/
Business Focused Articles About Ontology and AI
35. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved. 36
Publication Topic URL
CustomerThink High Fidelity Journey Models https://customerthink.com/improving-the-digital-experience-6-steps-to-
create-a-high-fidelity-journey-map/
Analytics Magazine Pandemic provides opportunity
to strengthen enterprises' tech
infrastructure
https://pubsonline.informs.org/do/10.1287/LYTX.2020.03.03/full/
Artificially Intelligent Podcast https://artificiallyintelligent.libsyn.com/102-the-ai-powered-enterprise-
with-seth-earley
TDWI Upside Excerpt printed https://tdwi.org/articles/2020/03/17/adv-all-ai-powered-future.aspx
Information Week AI Hot Spots: Where Is Artificial
Intelligence Heading Now?
https://www.informationweek.com/big-data/ai-machine-learning/ai-hot-
spots-where-is-artificial-intelligence-heading-now/d/d-
id/1337237?page_number=1
The Enterprisers
Project
8 reasons AI projects fail https://enterprisersproject.com/article/2020/3/why-ai-projects-fail-8-
reasons
Big Data Quarterly Harnessing the Power of AI for
the Enterprise: Q&A with Seth
Earley
http://www.dbta.com/BigDataQuarterly/Articles/Harnessing-the-Power-
of-AI-for-the-Enterprise-139509.aspx
E-Commerce Times The Architectural Imperative for
AI-Powered E-Commerce
https://www.ecommercetimes.com/story/86530.html
Business Focused Articles About Ontology and AI
36. Copyright © 2020 Earley Information Science, Inc. All Rights Reserved.
IEEE IT Computing Edge articles:
“There’s No AI without IA”
“The Problem with AI”
www.earley.com
Follow me on Twitter: @sethearley
37
Seth Earley
CEO – Earley Information Science
AUTHOR – The AI-Powered Enterprise
________________________________________________
Cell: 781-820-8080
Email: seth@earley.com
Web: www.earley.com
Connect with me on LinkedIn:
https://www.linkedin.com/in/sethearley
And on Facebook:
https://www.facebook.com/seth.earley