SlideShare a Scribd company logo
Codename Nancy
AI: lessons from the National
Library of Norway
AI @ the National Library of Norway
Svein Arne Brygfjeld
NB AILAB
From «cataloging rules» to «approximate, but
good enough»
Bsckground: The National what…?
• The National Library of Norway
• Major Norwegian memory institution under the Ministry of Culture
• All information carriers, also historical AV
• Act on legal deposit
• ~500 employees
• Two sites 1000 km apart, Oslo and Mo i Rana
Today
• Motivation, Nancy & the National Library of Norway
• Short introduction to Machine Learning
• What can AI do for libraries?
• What can libraries do for AI?
• Community and Conclusions
Why AI now?
• The future doesn’t come, we create it
• And it is possible
• Machines, software, content and metadata, human resources
• And we are libraries…
DATA
INFORMATION
KNOWLEDGE
WISDOM
Low-hanging fruit
• Open/available training and test sets
• Imagenet and others
• Digital library collections
• Open source democratized software
• TensorFlow and others
• Commercial services
• API’s for speech-to-text, still image and moving image analyzes, natural language
processing and more
Nancy
• Umbrella for AI innovations and applications at NLN
• Defined activity, NB AILAB, reporting directly to the
national librarian
• Input to conversation about AI through small scale
projects and experiments, typically weeks to months
in volume
• Currently 3-4 people on full time + some sourcing
Our digital shift – 2005/2006
• Status 2005
• What has happened?
• Just one thing. It is called «Internet» - all information is expected
to be found there
• 2006
• Digitize the collection
• Make the digital collection available always and anywhere
Our digital shift – today
• Massive digital collection with good metadata, like
• Text: Books, journals, newspapers
• Audio: Radio broadcast, music, documentary
• Images/moving images, TV
• Heavy use by external users
Main machine learning profiles
• Unsupervised learning
• Supervised learning
• Reinforced learning
ML: From exact to approximate
• Software development
is
• Precise, predictable
• IF-THEN-ELSE
• «Right» or «wrong»
• Changing to
• Approximate performance
• Experience based learning
• Good enough
Supervised: Training, test, use
ML Platform
i.e. TensorFlow
Algorithm & Model
«Black box»
Content
Metadata
Training
Content
Result
i.e. metadata
classification etc
Use
MetadataContent
Test
NOT
good enough
Good enough
(measures!)
Supervised: Training, test, use
ML Platform
i.e. TensorFlow
Algorithm & Model
«Black box»
Content
Metadata
Training
Content
New catalog data
Use
MetadataContent
Test
Books
Catalog data
Unread
books
Correct
catalog data
New
books
Examples from our AI lab
AI for libraries? (1 of 7)
Simple classification
Simple classification
Grouping, description & more
• Litterature groups as an example
• Four groups for commercial use
• Each book belongs to one group only
• «Nancy, in which litterature group does this book belong?»
• She is right in approx 95% of the test cases
AI for libraries? (2 of 7)
Dewey Decimal Classification
Complex classification
• Dewey Decimal Classification
Dewey Decimal Classification
• Nancy, could you please classify this article by 3, 4, 5 and
6 digits Dewey?
• Norart (scientific database) as metadata
• Born digital content, artificial articles to improve training set
• Result: 70-92% performance (in some rare cases, 100%)
AI for libraries? (3 of 7)
Language corpus
Making a language corpus
Some sample text
+
ML based quality
control
=
Large high quality corpus for
spoken Norwegian
(platform for machine learning)
AI for libraries? (4 of 7)
Supporting/changing workflows
Supporting existing workflow
• Our sami bibliography misses works for the years 1988-1992
• Challenge: find digitized candidates
AI for libraries? (5 of 7)
Commercial services rather than in-house training
Commercial API’s
Machine Learning based services
- Alternative to in-house training
- Typically image classification, speech-to-text, video
analyzes, natural language processing
- General in terms of use, vocabulary etc
- Several vendors like Amazon, Clarify and Google
AI for libraries? (6 of 7)
A complete multimedia digital library
Nancy’s complex challenge
• Can we make a digital library based on machine learning only?
• The case
• Complete content for January 2011
• 250 newspapers (3.000 issues, 4.400.000 articles, 209.000 images)
• Two national radio network channels (742 hrs audio)
• One national TV channel (100 hrs video)
Metadata production based on
Machine Learning
• Persons (names)
• Places
• Organizations
• Time
• Relations
• Subject classification
News-
papers
Radio
TV
Design principle
Entity and
relation
extraction,
Subject
analyzes
Geo
location
Search
platform
Text
with
time
coding
Speech to
text
Video
analyzes
Text,
OCR,
objects
& more
Articles
and
images
Text and
image
extraction
News-
papers
Radio
TV
Geo (map) search
Person search example
AI for libraries? (7 of 7)
Finding similar objects
ML without learning
ML Plattform
TensorFlow
Algorithm & modell
«Black box» Content
Result
Grouping/clustering
Use
Content
Test
1. Similar photographs
2. Similar books
What can libraries do for AI?
• Open training and test sets
• Digital content with high quality metadata
• Access to development labs as alternative
• Domain expertise
• Measuring the performance of AI systems
• Pre-trained models
• Plug-in models for various types of content
Conclusions?
• We understand that we don’t understand, but that we need to
• Internal workflows may change to the better and more effective
• Our view on metadata may change radically
• Our collections may be more accessible
• We may contribute better to knowledge and understanding
• All-in-all: we may be better libraries
• International collaboration needed
• ai4lib, ai4lam, GoogleGroups
Announcement
NLN AND STANFORD LIBRARIES JOINT
FANTASTIC FUTURES 2. AI CONFERENCE
DEC 4-5, STANFORD UNIVERSITY PALO ALTO
Thank you
Svein Arne Brygfjeld
svein.arne.brygfjeld@nb.no

More Related Content

Similar to SCONUL Summer Conference 2019 - Svein Arne Brygfjeld

Introduction to NLP.pptx
Introduction to NLP.pptxIntroduction to NLP.pptx
Introduction to NLP.pptx
buivantan_uneti
 
Redesigning our Combine Harvester
Redesigning our Combine HarvesterRedesigning our Combine Harvester
Redesigning our Combine Harvester
Try PurpleSearch
 
Textkernel talks - introduction to Textkernel
Textkernel talks - introduction to TextkernelTextkernel talks - introduction to Textkernel
Textkernel talks - introduction to Textkernel
Textkernel
 
Natural language Analysis
Natural language AnalysisNatural language Analysis
Natural language Analysis
Rudradeb Mitra
 
AILABS - Lecture Series - Is AI the New Electricity? - Advances In Machine Le...
AILABS - Lecture Series - Is AI the New Electricity? - Advances In Machine Le...AILABS - Lecture Series - Is AI the New Electricity? - Advances In Machine Le...
AILABS - Lecture Series - Is AI the New Electricity? - Advances In Machine Le...
AILABS Academy
 
Semantics as a service at EMBL-EBI
Semantics as a service at EMBL-EBISemantics as a service at EMBL-EBI
Semantics as a service at EMBL-EBI
Simon Jupp
 
Introduction to natural language processing (NLP)
Introduction to natural language processing (NLP)Introduction to natural language processing (NLP)
Introduction to natural language processing (NLP)
Alia Hamwi
 
Transkribus | Günter Mühlberger
Transkribus | Günter MühlbergerTranskribus | Günter Mühlberger
Transkribus | Günter Mühlberger
Netwerk Oorlogsbronnen
 
DLCS
DLCSDLCS
DLCS
Tom Crane
 
Beyond the Symbols: A 30-minute Overview of NLP
Beyond the Symbols: A 30-minute Overview of NLPBeyond the Symbols: A 30-minute Overview of NLP
Beyond the Symbols: A 30-minute Overview of NLP
MENGSAYLOEM1
 
AI presentation and introduction - Retrieval Augmented Generation RAG 101
AI presentation and introduction - Retrieval Augmented Generation RAG 101AI presentation and introduction - Retrieval Augmented Generation RAG 101
AI presentation and introduction - Retrieval Augmented Generation RAG 101
vincent683379
 
My projects at University of Oxford e-Research Centre - Nov 2014
My projects at University of Oxford e-Research Centre - Nov 2014My projects at University of Oxford e-Research Centre - Nov 2014
My projects at University of Oxford e-Research Centre - Nov 2014
Susanna-Assunta Sansone
 
Natural language processing and search
Natural language processing and searchNatural language processing and search
Natural language processing and search
Nathan McMinn
 
Digital Humanities Clinics – Leading Dutch Librarians into DH. Lotte Wilms, N...
Digital Humanities Clinics – Leading Dutch Librarians into DH. Lotte Wilms, N...Digital Humanities Clinics – Leading Dutch Librarians into DH. Lotte Wilms, N...
Digital Humanities Clinics – Leading Dutch Librarians into DH. Lotte Wilms, N...
LIBER Europe
 
How Oracle Uses CrowdFlower For Sentiment Analysis
How Oracle Uses CrowdFlower For Sentiment AnalysisHow Oracle Uses CrowdFlower For Sentiment Analysis
How Oracle Uses CrowdFlower For Sentiment Analysis
CrowdFlower
 
Introduction to Text Mining
Introduction to Text MiningIntroduction to Text Mining
Introduction to Text Mining
Minha Hwang
 
Can Repositories be fun? Thinking about repositories
Can Repositories be fun? Thinking about repositoriesCan Repositories be fun? Thinking about repositories
Can Repositories be fun? Thinking about repositories
Patrick Danowski
 
COAR Next Generation Repositories WG - Text mining and Recommender system sto...
COAR Next Generation Repositories WG - Text mining and Recommender system sto...COAR Next Generation Repositories WG - Text mining and Recommender system sto...
COAR Next Generation Repositories WG - Text mining and Recommender system sto...
petrknoth
 
Ontologies for multimedia: the Semantic Culture Web
Ontologies for multimedia: the Semantic Culture WebOntologies for multimedia: the Semantic Culture Web
Ontologies for multimedia: the Semantic Culture Web
Guus Schreiber
 
CICLing 2016
CICLing 2016CICLing 2016

Similar to SCONUL Summer Conference 2019 - Svein Arne Brygfjeld (20)

Introduction to NLP.pptx
Introduction to NLP.pptxIntroduction to NLP.pptx
Introduction to NLP.pptx
 
Redesigning our Combine Harvester
Redesigning our Combine HarvesterRedesigning our Combine Harvester
Redesigning our Combine Harvester
 
Textkernel talks - introduction to Textkernel
Textkernel talks - introduction to TextkernelTextkernel talks - introduction to Textkernel
Textkernel talks - introduction to Textkernel
 
Natural language Analysis
Natural language AnalysisNatural language Analysis
Natural language Analysis
 
AILABS - Lecture Series - Is AI the New Electricity? - Advances In Machine Le...
AILABS - Lecture Series - Is AI the New Electricity? - Advances In Machine Le...AILABS - Lecture Series - Is AI the New Electricity? - Advances In Machine Le...
AILABS - Lecture Series - Is AI the New Electricity? - Advances In Machine Le...
 
Semantics as a service at EMBL-EBI
Semantics as a service at EMBL-EBISemantics as a service at EMBL-EBI
Semantics as a service at EMBL-EBI
 
Introduction to natural language processing (NLP)
Introduction to natural language processing (NLP)Introduction to natural language processing (NLP)
Introduction to natural language processing (NLP)
 
Transkribus | Günter Mühlberger
Transkribus | Günter MühlbergerTranskribus | Günter Mühlberger
Transkribus | Günter Mühlberger
 
DLCS
DLCSDLCS
DLCS
 
Beyond the Symbols: A 30-minute Overview of NLP
Beyond the Symbols: A 30-minute Overview of NLPBeyond the Symbols: A 30-minute Overview of NLP
Beyond the Symbols: A 30-minute Overview of NLP
 
AI presentation and introduction - Retrieval Augmented Generation RAG 101
AI presentation and introduction - Retrieval Augmented Generation RAG 101AI presentation and introduction - Retrieval Augmented Generation RAG 101
AI presentation and introduction - Retrieval Augmented Generation RAG 101
 
My projects at University of Oxford e-Research Centre - Nov 2014
My projects at University of Oxford e-Research Centre - Nov 2014My projects at University of Oxford e-Research Centre - Nov 2014
My projects at University of Oxford e-Research Centre - Nov 2014
 
Natural language processing and search
Natural language processing and searchNatural language processing and search
Natural language processing and search
 
Digital Humanities Clinics – Leading Dutch Librarians into DH. Lotte Wilms, N...
Digital Humanities Clinics – Leading Dutch Librarians into DH. Lotte Wilms, N...Digital Humanities Clinics – Leading Dutch Librarians into DH. Lotte Wilms, N...
Digital Humanities Clinics – Leading Dutch Librarians into DH. Lotte Wilms, N...
 
How Oracle Uses CrowdFlower For Sentiment Analysis
How Oracle Uses CrowdFlower For Sentiment AnalysisHow Oracle Uses CrowdFlower For Sentiment Analysis
How Oracle Uses CrowdFlower For Sentiment Analysis
 
Introduction to Text Mining
Introduction to Text MiningIntroduction to Text Mining
Introduction to Text Mining
 
Can Repositories be fun? Thinking about repositories
Can Repositories be fun? Thinking about repositoriesCan Repositories be fun? Thinking about repositories
Can Repositories be fun? Thinking about repositories
 
COAR Next Generation Repositories WG - Text mining and Recommender system sto...
COAR Next Generation Repositories WG - Text mining and Recommender system sto...COAR Next Generation Repositories WG - Text mining and Recommender system sto...
COAR Next Generation Repositories WG - Text mining and Recommender system sto...
 
Ontologies for multimedia: the Semantic Culture Web
Ontologies for multimedia: the Semantic Culture WebOntologies for multimedia: the Semantic Culture Web
Ontologies for multimedia: the Semantic Culture Web
 
CICLing 2016
CICLing 2016CICLing 2016
CICLing 2016
 

More from sconul

SCONUL Library Design Awards 2019 - Laura Norris
SCONUL Library Design Awards 2019 - Laura NorrisSCONUL Library Design Awards 2019 - Laura Norris
SCONUL Library Design Awards 2019 - Laura Norris
sconul
 
SCONUL Library Design Awards 2019 - Professor Nick petford
SCONUL Library Design Awards 2019 - Professor Nick petfordSCONUL Library Design Awards 2019 - Professor Nick petford
SCONUL Library Design Awards 2019 - Professor Nick petford
sconul
 
SCONUL Library Design Awards 2019 - University of Kent
SCONUL Library Design Awards 2019 - University of KentSCONUL Library Design Awards 2019 - University of Kent
SCONUL Library Design Awards 2019 - University of Kent
sconul
 
SCONUL Library Design Awards 2019 - University of Roehampton
SCONUL Library Design Awards 2019 - University of RoehamptonSCONUL Library Design Awards 2019 - University of Roehampton
SCONUL Library Design Awards 2019 - University of Roehampton
sconul
 
SCONUL Library Design Awards 2019 - Royal College of Surgeons in Ireland
SCONUL Library Design Awards 2019 - Royal College of Surgeons in IrelandSCONUL Library Design Awards 2019 - Royal College of Surgeons in Ireland
SCONUL Library Design Awards 2019 - Royal College of Surgeons in Ireland
sconul
 
SCONUL Library Design Awards 2019 - University of Leeds
SCONUL Library Design Awards 2019 - University of LeedsSCONUL Library Design Awards 2019 - University of Leeds
SCONUL Library Design Awards 2019 - University of Leeds
sconul
 
SCONUL Library Design Awards 2019 - University of Essex
SCONUL Library Design Awards 2019 - University of EssexSCONUL Library Design Awards 2019 - University of Essex
SCONUL Library Design Awards 2019 - University of Essex
sconul
 
SCONUL Library Design Awards 2019 - University of Birmingham
SCONUL Library Design Awards 2019 - University of BirminghamSCONUL Library Design Awards 2019 - University of Birmingham
SCONUL Library Design Awards 2019 - University of Birmingham
sconul
 
SCONUL Summer Conference 2019 - Dr Tamsin Burland
SCONUL Summer Conference 2019 - Dr Tamsin BurlandSCONUL Summer Conference 2019 - Dr Tamsin Burland
SCONUL Summer Conference 2019 - Dr Tamsin Burland
sconul
 
SCONUL Summer Conference 2019 - Merrilee Proffitt
SCONUL Summer Conference 2019 - Merrilee ProffittSCONUL Summer Conference 2019 - Merrilee Proffitt
SCONUL Summer Conference 2019 - Merrilee Proffitt
sconul
 
SCONUL Summer Conference 2019 - David Sweeney
SCONUL Summer Conference 2019 - David SweeneySCONUL Summer Conference 2019 - David Sweeney
SCONUL Summer Conference 2019 - David Sweeney
sconul
 
SCONUL Summer Conference 2019 - Alison Selina & Suzi Robinson
SCONUL Summer Conference 2019 - Alison Selina & Suzi RobinsonSCONUL Summer Conference 2019 - Alison Selina & Suzi Robinson
SCONUL Summer Conference 2019 - Alison Selina & Suzi Robinson
sconul
 
SCONUL Summer Conference 2019 - Regina Everitt, Caroline Taylor and Dr Mohamm...
SCONUL Summer Conference 2019 - Regina Everitt, Caroline Taylor and Dr Mohamm...SCONUL Summer Conference 2019 - Regina Everitt, Caroline Taylor and Dr Mohamm...
SCONUL Summer Conference 2019 - Regina Everitt, Caroline Taylor and Dr Mohamm...
sconul
 
SCONUL Summer Conference 2019 - Liz Waller & Nick Barratt
SCONUL Summer Conference 2019 - Liz Waller & Nick BarrattSCONUL Summer Conference 2019 - Liz Waller & Nick Barratt
SCONUL Summer Conference 2019 - Liz Waller & Nick Barratt
sconul
 
SCONUL Summer Conference 2019 - Lidia Borrell-Damián
SCONUL Summer Conference 2019 - Lidia Borrell-DamiánSCONUL Summer Conference 2019 - Lidia Borrell-Damián
SCONUL Summer Conference 2019 - Lidia Borrell-Damián
sconul
 
SCONUL Summer Conference 2018 - Nicole coleman
SCONUL Summer Conference 2018 - Nicole colemanSCONUL Summer Conference 2018 - Nicole coleman
SCONUL Summer Conference 2018 - Nicole coleman
sconul
 
SCONUL Summer Conference 2018 - Simon Walker
SCONUL Summer Conference 2018 - Simon WalkerSCONUL Summer Conference 2018 - Simon Walker
SCONUL Summer Conference 2018 - Simon Walker
sconul
 
SCONUL Summer Conference - 2018 - Rufus Pollock
SCONUL Summer Conference - 2018 - Rufus PollockSCONUL Summer Conference - 2018 - Rufus Pollock
SCONUL Summer Conference - 2018 - Rufus Pollock
sconul
 
SCONUL Summer Conference 2018 - Richard Watson
SCONUL Summer Conference 2018 - Richard WatsonSCONUL Summer Conference 2018 - Richard Watson
SCONUL Summer Conference 2018 - Richard Watson
sconul
 
SCONUL Summer Conference 2018 - Paul Feldman
SCONUL Summer Conference 2018 - Paul FeldmanSCONUL Summer Conference 2018 - Paul Feldman
SCONUL Summer Conference 2018 - Paul Feldman
sconul
 

More from sconul (20)

SCONUL Library Design Awards 2019 - Laura Norris
SCONUL Library Design Awards 2019 - Laura NorrisSCONUL Library Design Awards 2019 - Laura Norris
SCONUL Library Design Awards 2019 - Laura Norris
 
SCONUL Library Design Awards 2019 - Professor Nick petford
SCONUL Library Design Awards 2019 - Professor Nick petfordSCONUL Library Design Awards 2019 - Professor Nick petford
SCONUL Library Design Awards 2019 - Professor Nick petford
 
SCONUL Library Design Awards 2019 - University of Kent
SCONUL Library Design Awards 2019 - University of KentSCONUL Library Design Awards 2019 - University of Kent
SCONUL Library Design Awards 2019 - University of Kent
 
SCONUL Library Design Awards 2019 - University of Roehampton
SCONUL Library Design Awards 2019 - University of RoehamptonSCONUL Library Design Awards 2019 - University of Roehampton
SCONUL Library Design Awards 2019 - University of Roehampton
 
SCONUL Library Design Awards 2019 - Royal College of Surgeons in Ireland
SCONUL Library Design Awards 2019 - Royal College of Surgeons in IrelandSCONUL Library Design Awards 2019 - Royal College of Surgeons in Ireland
SCONUL Library Design Awards 2019 - Royal College of Surgeons in Ireland
 
SCONUL Library Design Awards 2019 - University of Leeds
SCONUL Library Design Awards 2019 - University of LeedsSCONUL Library Design Awards 2019 - University of Leeds
SCONUL Library Design Awards 2019 - University of Leeds
 
SCONUL Library Design Awards 2019 - University of Essex
SCONUL Library Design Awards 2019 - University of EssexSCONUL Library Design Awards 2019 - University of Essex
SCONUL Library Design Awards 2019 - University of Essex
 
SCONUL Library Design Awards 2019 - University of Birmingham
SCONUL Library Design Awards 2019 - University of BirminghamSCONUL Library Design Awards 2019 - University of Birmingham
SCONUL Library Design Awards 2019 - University of Birmingham
 
SCONUL Summer Conference 2019 - Dr Tamsin Burland
SCONUL Summer Conference 2019 - Dr Tamsin BurlandSCONUL Summer Conference 2019 - Dr Tamsin Burland
SCONUL Summer Conference 2019 - Dr Tamsin Burland
 
SCONUL Summer Conference 2019 - Merrilee Proffitt
SCONUL Summer Conference 2019 - Merrilee ProffittSCONUL Summer Conference 2019 - Merrilee Proffitt
SCONUL Summer Conference 2019 - Merrilee Proffitt
 
SCONUL Summer Conference 2019 - David Sweeney
SCONUL Summer Conference 2019 - David SweeneySCONUL Summer Conference 2019 - David Sweeney
SCONUL Summer Conference 2019 - David Sweeney
 
SCONUL Summer Conference 2019 - Alison Selina & Suzi Robinson
SCONUL Summer Conference 2019 - Alison Selina & Suzi RobinsonSCONUL Summer Conference 2019 - Alison Selina & Suzi Robinson
SCONUL Summer Conference 2019 - Alison Selina & Suzi Robinson
 
SCONUL Summer Conference 2019 - Regina Everitt, Caroline Taylor and Dr Mohamm...
SCONUL Summer Conference 2019 - Regina Everitt, Caroline Taylor and Dr Mohamm...SCONUL Summer Conference 2019 - Regina Everitt, Caroline Taylor and Dr Mohamm...
SCONUL Summer Conference 2019 - Regina Everitt, Caroline Taylor and Dr Mohamm...
 
SCONUL Summer Conference 2019 - Liz Waller & Nick Barratt
SCONUL Summer Conference 2019 - Liz Waller & Nick BarrattSCONUL Summer Conference 2019 - Liz Waller & Nick Barratt
SCONUL Summer Conference 2019 - Liz Waller & Nick Barratt
 
SCONUL Summer Conference 2019 - Lidia Borrell-Damián
SCONUL Summer Conference 2019 - Lidia Borrell-DamiánSCONUL Summer Conference 2019 - Lidia Borrell-Damián
SCONUL Summer Conference 2019 - Lidia Borrell-Damián
 
SCONUL Summer Conference 2018 - Nicole coleman
SCONUL Summer Conference 2018 - Nicole colemanSCONUL Summer Conference 2018 - Nicole coleman
SCONUL Summer Conference 2018 - Nicole coleman
 
SCONUL Summer Conference 2018 - Simon Walker
SCONUL Summer Conference 2018 - Simon WalkerSCONUL Summer Conference 2018 - Simon Walker
SCONUL Summer Conference 2018 - Simon Walker
 
SCONUL Summer Conference - 2018 - Rufus Pollock
SCONUL Summer Conference - 2018 - Rufus PollockSCONUL Summer Conference - 2018 - Rufus Pollock
SCONUL Summer Conference - 2018 - Rufus Pollock
 
SCONUL Summer Conference 2018 - Richard Watson
SCONUL Summer Conference 2018 - Richard WatsonSCONUL Summer Conference 2018 - Richard Watson
SCONUL Summer Conference 2018 - Richard Watson
 
SCONUL Summer Conference 2018 - Paul Feldman
SCONUL Summer Conference 2018 - Paul FeldmanSCONUL Summer Conference 2018 - Paul Feldman
SCONUL Summer Conference 2018 - Paul Feldman
 

Recently uploaded

NEWSPAPERS - QUESTION 1 - REVISION POWERPOINT.pptx
NEWSPAPERS - QUESTION 1 - REVISION POWERPOINT.pptxNEWSPAPERS - QUESTION 1 - REVISION POWERPOINT.pptx
NEWSPAPERS - QUESTION 1 - REVISION POWERPOINT.pptx
iammrhaywood
 
How to Setup Warehouse & Location in Odoo 17 Inventory
How to Setup Warehouse & Location in Odoo 17 InventoryHow to Setup Warehouse & Location in Odoo 17 Inventory
How to Setup Warehouse & Location in Odoo 17 Inventory
Celine George
 
Leveraging Generative AI to Drive Nonprofit Innovation
Leveraging Generative AI to Drive Nonprofit InnovationLeveraging Generative AI to Drive Nonprofit Innovation
Leveraging Generative AI to Drive Nonprofit Innovation
TechSoup
 
Electric Fetus - Record Store Scavenger Hunt
Electric Fetus - Record Store Scavenger HuntElectric Fetus - Record Store Scavenger Hunt
Electric Fetus - Record Store Scavenger Hunt
RamseyBerglund
 
What is Digital Literacy? A guest blog from Andy McLaughlin, University of Ab...
What is Digital Literacy? A guest blog from Andy McLaughlin, University of Ab...What is Digital Literacy? A guest blog from Andy McLaughlin, University of Ab...
What is Digital Literacy? A guest blog from Andy McLaughlin, University of Ab...
GeorgeMilliken2
 
BÀI TẬP BỔ TRỢ TIẾNG ANH LỚP 9 CẢ NĂM - GLOBAL SUCCESS - NĂM HỌC 2024-2025 - ...
BÀI TẬP BỔ TRỢ TIẾNG ANH LỚP 9 CẢ NĂM - GLOBAL SUCCESS - NĂM HỌC 2024-2025 - ...BÀI TẬP BỔ TRỢ TIẾNG ANH LỚP 9 CẢ NĂM - GLOBAL SUCCESS - NĂM HỌC 2024-2025 - ...
BÀI TẬP BỔ TRỢ TIẾNG ANH LỚP 9 CẢ NĂM - GLOBAL SUCCESS - NĂM HỌC 2024-2025 - ...
Nguyen Thanh Tu Collection
 
Level 3 NCEA - NZ: A Nation In the Making 1872 - 1900 SML.ppt
Level 3 NCEA - NZ: A  Nation In the Making 1872 - 1900 SML.pptLevel 3 NCEA - NZ: A  Nation In the Making 1872 - 1900 SML.ppt
Level 3 NCEA - NZ: A Nation In the Making 1872 - 1900 SML.ppt
Henry Hollis
 
مصحف القراءات العشر أعد أحرف الخلاف سمير بسيوني.pdf
مصحف القراءات العشر   أعد أحرف الخلاف سمير بسيوني.pdfمصحف القراءات العشر   أعد أحرف الخلاف سمير بسيوني.pdf
مصحف القراءات العشر أعد أحرف الخلاف سمير بسيوني.pdf
سمير بسيوني
 
B. Ed Syllabus for babasaheb ambedkar education university.pdf
B. Ed Syllabus for babasaheb ambedkar education university.pdfB. Ed Syllabus for babasaheb ambedkar education university.pdf
B. Ed Syllabus for babasaheb ambedkar education university.pdf
BoudhayanBhattachari
 
writing about opinions about Australia the movie
writing about opinions about Australia the moviewriting about opinions about Australia the movie
writing about opinions about Australia the movie
Nicholas Montgomery
 
BÀI TẬP BỔ TRỢ TIẾNG ANH 8 CẢ NĂM - GLOBAL SUCCESS - NĂM HỌC 2023-2024 (CÓ FI...
BÀI TẬP BỔ TRỢ TIẾNG ANH 8 CẢ NĂM - GLOBAL SUCCESS - NĂM HỌC 2023-2024 (CÓ FI...BÀI TẬP BỔ TRỢ TIẾNG ANH 8 CẢ NĂM - GLOBAL SUCCESS - NĂM HỌC 2023-2024 (CÓ FI...
BÀI TẬP BỔ TRỢ TIẾNG ANH 8 CẢ NĂM - GLOBAL SUCCESS - NĂM HỌC 2023-2024 (CÓ FI...
Nguyen Thanh Tu Collection
 
Pharmaceutics Pharmaceuticals best of brub
Pharmaceutics Pharmaceuticals best of brubPharmaceutics Pharmaceuticals best of brub
Pharmaceutics Pharmaceuticals best of brub
danielkiash986
 
math operations ued in python and all used
math operations ued in python and all usedmath operations ued in python and all used
math operations ued in python and all used
ssuser13ffe4
 
Présentationvvvvvvvvvvvvvvvvvvvvvvvvvvvv2.pptx
Présentationvvvvvvvvvvvvvvvvvvvvvvvvvvvv2.pptxPrésentationvvvvvvvvvvvvvvvvvvvvvvvvvvvv2.pptx
Présentationvvvvvvvvvvvvvvvvvvvvvvvvvvvv2.pptx
siemaillard
 
C1 Rubenstein AP HuG xxxxxxxxxxxxxx.pptx
C1 Rubenstein AP HuG xxxxxxxxxxxxxx.pptxC1 Rubenstein AP HuG xxxxxxxxxxxxxx.pptx
C1 Rubenstein AP HuG xxxxxxxxxxxxxx.pptx
mulvey2
 
SWOT analysis in the project Keeping the Memory @live.pptx
SWOT analysis in the project Keeping the Memory @live.pptxSWOT analysis in the project Keeping the Memory @live.pptx
SWOT analysis in the project Keeping the Memory @live.pptx
zuzanka
 
Mule event processing models | MuleSoft Mysore Meetup #47
Mule event processing models | MuleSoft Mysore Meetup #47Mule event processing models | MuleSoft Mysore Meetup #47
Mule event processing models | MuleSoft Mysore Meetup #47
MysoreMuleSoftMeetup
 
RESULTS OF THE EVALUATION QUESTIONNAIRE.pptx
RESULTS OF THE EVALUATION QUESTIONNAIRE.pptxRESULTS OF THE EVALUATION QUESTIONNAIRE.pptx
RESULTS OF THE EVALUATION QUESTIONNAIRE.pptx
zuzanka
 
Film vocab for eal 3 students: Australia the movie
Film vocab for eal 3 students: Australia the movieFilm vocab for eal 3 students: Australia the movie
Film vocab for eal 3 students: Australia the movie
Nicholas Montgomery
 
A Independência da América Espanhola LAPBOOK.pdf
A Independência da América Espanhola LAPBOOK.pdfA Independência da América Espanhola LAPBOOK.pdf
A Independência da América Espanhola LAPBOOK.pdf
Jean Carlos Nunes Paixão
 

Recently uploaded (20)

NEWSPAPERS - QUESTION 1 - REVISION POWERPOINT.pptx
NEWSPAPERS - QUESTION 1 - REVISION POWERPOINT.pptxNEWSPAPERS - QUESTION 1 - REVISION POWERPOINT.pptx
NEWSPAPERS - QUESTION 1 - REVISION POWERPOINT.pptx
 
How to Setup Warehouse & Location in Odoo 17 Inventory
How to Setup Warehouse & Location in Odoo 17 InventoryHow to Setup Warehouse & Location in Odoo 17 Inventory
How to Setup Warehouse & Location in Odoo 17 Inventory
 
Leveraging Generative AI to Drive Nonprofit Innovation
Leveraging Generative AI to Drive Nonprofit InnovationLeveraging Generative AI to Drive Nonprofit Innovation
Leveraging Generative AI to Drive Nonprofit Innovation
 
Electric Fetus - Record Store Scavenger Hunt
Electric Fetus - Record Store Scavenger HuntElectric Fetus - Record Store Scavenger Hunt
Electric Fetus - Record Store Scavenger Hunt
 
What is Digital Literacy? A guest blog from Andy McLaughlin, University of Ab...
What is Digital Literacy? A guest blog from Andy McLaughlin, University of Ab...What is Digital Literacy? A guest blog from Andy McLaughlin, University of Ab...
What is Digital Literacy? A guest blog from Andy McLaughlin, University of Ab...
 
BÀI TẬP BỔ TRỢ TIẾNG ANH LỚP 9 CẢ NĂM - GLOBAL SUCCESS - NĂM HỌC 2024-2025 - ...
BÀI TẬP BỔ TRỢ TIẾNG ANH LỚP 9 CẢ NĂM - GLOBAL SUCCESS - NĂM HỌC 2024-2025 - ...BÀI TẬP BỔ TRỢ TIẾNG ANH LỚP 9 CẢ NĂM - GLOBAL SUCCESS - NĂM HỌC 2024-2025 - ...
BÀI TẬP BỔ TRỢ TIẾNG ANH LỚP 9 CẢ NĂM - GLOBAL SUCCESS - NĂM HỌC 2024-2025 - ...
 
Level 3 NCEA - NZ: A Nation In the Making 1872 - 1900 SML.ppt
Level 3 NCEA - NZ: A  Nation In the Making 1872 - 1900 SML.pptLevel 3 NCEA - NZ: A  Nation In the Making 1872 - 1900 SML.ppt
Level 3 NCEA - NZ: A Nation In the Making 1872 - 1900 SML.ppt
 
مصحف القراءات العشر أعد أحرف الخلاف سمير بسيوني.pdf
مصحف القراءات العشر   أعد أحرف الخلاف سمير بسيوني.pdfمصحف القراءات العشر   أعد أحرف الخلاف سمير بسيوني.pdf
مصحف القراءات العشر أعد أحرف الخلاف سمير بسيوني.pdf
 
B. Ed Syllabus for babasaheb ambedkar education university.pdf
B. Ed Syllabus for babasaheb ambedkar education university.pdfB. Ed Syllabus for babasaheb ambedkar education university.pdf
B. Ed Syllabus for babasaheb ambedkar education university.pdf
 
writing about opinions about Australia the movie
writing about opinions about Australia the moviewriting about opinions about Australia the movie
writing about opinions about Australia the movie
 
BÀI TẬP BỔ TRỢ TIẾNG ANH 8 CẢ NĂM - GLOBAL SUCCESS - NĂM HỌC 2023-2024 (CÓ FI...
BÀI TẬP BỔ TRỢ TIẾNG ANH 8 CẢ NĂM - GLOBAL SUCCESS - NĂM HỌC 2023-2024 (CÓ FI...BÀI TẬP BỔ TRỢ TIẾNG ANH 8 CẢ NĂM - GLOBAL SUCCESS - NĂM HỌC 2023-2024 (CÓ FI...
BÀI TẬP BỔ TRỢ TIẾNG ANH 8 CẢ NĂM - GLOBAL SUCCESS - NĂM HỌC 2023-2024 (CÓ FI...
 
Pharmaceutics Pharmaceuticals best of brub
Pharmaceutics Pharmaceuticals best of brubPharmaceutics Pharmaceuticals best of brub
Pharmaceutics Pharmaceuticals best of brub
 
math operations ued in python and all used
math operations ued in python and all usedmath operations ued in python and all used
math operations ued in python and all used
 
Présentationvvvvvvvvvvvvvvvvvvvvvvvvvvvv2.pptx
Présentationvvvvvvvvvvvvvvvvvvvvvvvvvvvv2.pptxPrésentationvvvvvvvvvvvvvvvvvvvvvvvvvvvv2.pptx
Présentationvvvvvvvvvvvvvvvvvvvvvvvvvvvv2.pptx
 
C1 Rubenstein AP HuG xxxxxxxxxxxxxx.pptx
C1 Rubenstein AP HuG xxxxxxxxxxxxxx.pptxC1 Rubenstein AP HuG xxxxxxxxxxxxxx.pptx
C1 Rubenstein AP HuG xxxxxxxxxxxxxx.pptx
 
SWOT analysis in the project Keeping the Memory @live.pptx
SWOT analysis in the project Keeping the Memory @live.pptxSWOT analysis in the project Keeping the Memory @live.pptx
SWOT analysis in the project Keeping the Memory @live.pptx
 
Mule event processing models | MuleSoft Mysore Meetup #47
Mule event processing models | MuleSoft Mysore Meetup #47Mule event processing models | MuleSoft Mysore Meetup #47
Mule event processing models | MuleSoft Mysore Meetup #47
 
RESULTS OF THE EVALUATION QUESTIONNAIRE.pptx
RESULTS OF THE EVALUATION QUESTIONNAIRE.pptxRESULTS OF THE EVALUATION QUESTIONNAIRE.pptx
RESULTS OF THE EVALUATION QUESTIONNAIRE.pptx
 
Film vocab for eal 3 students: Australia the movie
Film vocab for eal 3 students: Australia the movieFilm vocab for eal 3 students: Australia the movie
Film vocab for eal 3 students: Australia the movie
 
A Independência da América Espanhola LAPBOOK.pdf
A Independência da América Espanhola LAPBOOK.pdfA Independência da América Espanhola LAPBOOK.pdf
A Independência da América Espanhola LAPBOOK.pdf
 

SCONUL Summer Conference 2019 - Svein Arne Brygfjeld

  • 1. Codename Nancy AI: lessons from the National Library of Norway AI @ the National Library of Norway Svein Arne Brygfjeld NB AILAB
  • 2. From «cataloging rules» to «approximate, but good enough»
  • 3. Bsckground: The National what…? • The National Library of Norway • Major Norwegian memory institution under the Ministry of Culture • All information carriers, also historical AV • Act on legal deposit • ~500 employees • Two sites 1000 km apart, Oslo and Mo i Rana
  • 4. Today • Motivation, Nancy & the National Library of Norway • Short introduction to Machine Learning • What can AI do for libraries? • What can libraries do for AI? • Community and Conclusions
  • 5. Why AI now? • The future doesn’t come, we create it • And it is possible • Machines, software, content and metadata, human resources • And we are libraries… DATA INFORMATION KNOWLEDGE WISDOM
  • 6. Low-hanging fruit • Open/available training and test sets • Imagenet and others • Digital library collections • Open source democratized software • TensorFlow and others • Commercial services • API’s for speech-to-text, still image and moving image analyzes, natural language processing and more
  • 7. Nancy • Umbrella for AI innovations and applications at NLN • Defined activity, NB AILAB, reporting directly to the national librarian • Input to conversation about AI through small scale projects and experiments, typically weeks to months in volume • Currently 3-4 people on full time + some sourcing
  • 8. Our digital shift – 2005/2006 • Status 2005 • What has happened? • Just one thing. It is called «Internet» - all information is expected to be found there • 2006 • Digitize the collection • Make the digital collection available always and anywhere
  • 9. Our digital shift – today • Massive digital collection with good metadata, like • Text: Books, journals, newspapers • Audio: Radio broadcast, music, documentary • Images/moving images, TV • Heavy use by external users
  • 10. Main machine learning profiles • Unsupervised learning • Supervised learning • Reinforced learning
  • 11. ML: From exact to approximate • Software development is • Precise, predictable • IF-THEN-ELSE • «Right» or «wrong» • Changing to • Approximate performance • Experience based learning • Good enough
  • 12. Supervised: Training, test, use ML Platform i.e. TensorFlow Algorithm & Model «Black box» Content Metadata Training Content Result i.e. metadata classification etc Use MetadataContent Test NOT good enough Good enough (measures!)
  • 13. Supervised: Training, test, use ML Platform i.e. TensorFlow Algorithm & Model «Black box» Content Metadata Training Content New catalog data Use MetadataContent Test Books Catalog data Unread books Correct catalog data New books
  • 15. AI for libraries? (1 of 7) Simple classification
  • 16. Simple classification Grouping, description & more • Litterature groups as an example • Four groups for commercial use • Each book belongs to one group only • «Nancy, in which litterature group does this book belong?» • She is right in approx 95% of the test cases
  • 17. AI for libraries? (2 of 7) Dewey Decimal Classification
  • 18. Complex classification • Dewey Decimal Classification
  • 19. Dewey Decimal Classification • Nancy, could you please classify this article by 3, 4, 5 and 6 digits Dewey? • Norart (scientific database) as metadata • Born digital content, artificial articles to improve training set • Result: 70-92% performance (in some rare cases, 100%)
  • 20. AI for libraries? (3 of 7) Language corpus
  • 21. Making a language corpus Some sample text + ML based quality control = Large high quality corpus for spoken Norwegian (platform for machine learning)
  • 22. AI for libraries? (4 of 7) Supporting/changing workflows
  • 23. Supporting existing workflow • Our sami bibliography misses works for the years 1988-1992 • Challenge: find digitized candidates
  • 24. AI for libraries? (5 of 7) Commercial services rather than in-house training
  • 25. Commercial API’s Machine Learning based services - Alternative to in-house training - Typically image classification, speech-to-text, video analyzes, natural language processing - General in terms of use, vocabulary etc - Several vendors like Amazon, Clarify and Google
  • 26.
  • 27.
  • 28.
  • 29.
  • 30.
  • 31.
  • 32.
  • 33.
  • 34. AI for libraries? (6 of 7) A complete multimedia digital library
  • 35. Nancy’s complex challenge • Can we make a digital library based on machine learning only? • The case • Complete content for January 2011 • 250 newspapers (3.000 issues, 4.400.000 articles, 209.000 images) • Two national radio network channels (742 hrs audio) • One national TV channel (100 hrs video)
  • 36. Metadata production based on Machine Learning • Persons (names) • Places • Organizations • Time • Relations • Subject classification News- papers Radio TV
  • 37. Design principle Entity and relation extraction, Subject analyzes Geo location Search platform Text with time coding Speech to text Video analyzes Text, OCR, objects & more Articles and images Text and image extraction News- papers Radio TV
  • 40. AI for libraries? (7 of 7) Finding similar objects
  • 41. ML without learning ML Plattform TensorFlow Algorithm & modell «Black box» Content Result Grouping/clustering Use Content Test
  • 44. What can libraries do for AI? • Open training and test sets • Digital content with high quality metadata • Access to development labs as alternative • Domain expertise • Measuring the performance of AI systems • Pre-trained models • Plug-in models for various types of content
  • 45. Conclusions? • We understand that we don’t understand, but that we need to • Internal workflows may change to the better and more effective • Our view on metadata may change radically • Our collections may be more accessible • We may contribute better to knowledge and understanding • All-in-all: we may be better libraries
  • 46. • International collaboration needed • ai4lib, ai4lam, GoogleGroups
  • 47. Announcement NLN AND STANFORD LIBRARIES JOINT FANTASTIC FUTURES 2. AI CONFERENCE DEC 4-5, STANFORD UNIVERSITY PALO ALTO
  • 48. Thank you Svein Arne Brygfjeld svein.arne.brygfjeld@nb.no

Editor's Notes

  1. Auguste Rodin Il Penseroso, Michelangelo & Lorenzo di Medici