BISG WEBCAST -- Identification & Digital Publications

This BISG WEBCAST took placeTuesday, September 15, 2009 at 11:00 a.m. EDT.

B t h kTo register for future BISG Webcasts, please visit:http://www.bisg.org/event-cat-6-webcasts.php

Special thanks to our Webcast Sponsor:

Be sure to check your system prior to

the event!Visit the link below. If the li k i li k blSpecial thanks to our Webcast Sponsor:

The U.S. ISBN Agencylink is not clickable, copy and paste it into a new

browser window. Please take any actions suggested by the

systems check.

1SYSTEMS CHECK URL

http://events.meetingbridge.com/join/CheckOptions.aspx?rv=y&go=y

“Working to create a moreWorking to create a moreinformed, empowered and efficient

book industry supply chainbook industry supply chainfor both physical and digital products.”

Be sure to check your system prior toyour system prior to

the event!Visit the link to the right. If

the link is not clickable, copy and paste it into a new

browser window. Please take

2

any actions suggested by the systems check. SYSTEMS CHECK URL

http://events.meetingbridge.com/join/CheckOptions.aspx?rv=y&go=y

A d W i bAndy Weissberg

VP of Identifier Services &VP of Identifier Services &Corporate MarketingR.R. BowkerR.R. Bowker

Board of DirectorsInternational ISBN Agency

SYSTEMS CHECK URL

3

SYSTEMS CHECK URLhttp://events.meetingbridge.com/join/CheckOptions.aspx?rv=y&go=y

a quick audience poll...

Do you believe that digital manifestationsDo you believe that digital manifestations (specific formats, e.g., PDF, .Mobi) of books and/or other content should to be identified

with separate ISBNs?

Do you believe that digital manifestations (specific formats,e.g., PDF, .Mobi) of books and/or other content should to

be identified with separate ISBNs?

40 00%

50.00%

60.00%

Yes52.1%(76 )

10 00%

20.00%

30.00%

40.00% (76 votes)

No22.6%(33 votes)

Not sure

0.00%

10.00%

Yes52.1%

(76 votes)

No22.6%

(33 votes)

Not sure25.3%

(37 votes)

Not sure25.3%(37 votes)

( ) ( ) ( )

Results gathered during a live participant poll.

Mark Bide, Executive Director, EDItEURMark Bide is Executive Director of EDItEUR, the global trade standards organization for the book and serial publishing industries. He is also the Project Director for the ACAP Project, and a Director of Rightscom, the specialist media consultancy. Since the early 1990s, Mark has been closely involved in media standardization strategies and in the design and

f d d f id ifi i d d i h di H h k d imanagement of standards for identification and metadata in the media. He has worked in and around the publishing industry for nearly 40 years, having been a Director of the European subsidiaries of both CBS Publishing and John Wiley & Sons. He is a Visiting Professor of the University of the Arts London.

7

A brief introduction...

Identifiers are “just” a special class of name◦ Unique within a given context◦ Unique within a given context

Why do we assign identifiers?◦ Collocation – to bring together instances of the same thing◦ Disambiguation – to distinguish things that are not the same◦ Disambiguation to distinguish things that are not the same

What does “the same” mean?◦ Whether things are or are not the same is always contextual◦ For example an ISBN identifies instances as being “the same”For example, an ISBN identifies instances as being the same

for particular purposes – the meaning is not universal Why does this matter?◦ Unambiguous communication…g◦ …particularly from machine to machine (people don’t often

use unique identifiers in discourse – “that one over there” is usually enough)

When there is a need to communicate across organizational boundaries – within a supply chain…organizational boundaries within a supply chain…

…particularly where anyone in the supply chain needs to manage and aggregate information from multiple sources

h l l l d l l◦ That means nearly everyone, particularly in a digital supply chain

What matters about standard identifiers?◦ That their semantic should be clear to everyoneThat their semantic should be clear to everyone…◦ …in other words, everyone in the chain knows what type of

thing they are identifying So, an ISBN identifies a book, right?

W ll◦ Well, no◦ It identifies a product in the book supply chain…but we have

tried to make the standard do so much more

Some quotes from the PersonaNonData blog(Michael Cairns) August 4th(Michael Cairns) August 4th

ISBN is DeadI am increasingly concerned about the future health of the ISBN. In its current form the ISBN is not yet dead but therein lies the problem: ‘in its current form.’…As a community, we need to recognize that the ISBN

b d d k d d hmay not be meeting its intended market need and that the future may make this deficiency even more stark. …Into this mix I would also add that ISBN can no longer

d ll i d d f h id ifi hstand generally independent of other identifiers, such as a work ID or party ID.

Identity is the critical item of metadata for i biliinteroperability…◦ …“Are we talking about the same thing”?

The book industry recognised this very early –y g y yusing the ISBN to identify products…◦ …but then unfortunately went on to use the ISBN to identify

everything elseeverything else…◦ …and built systems that were entirely ISBN-centric (a folly

to which will return)

The “Master ISBN” is a commonly used as a proxy “work” identifier in publishers’ systemsidentifier in publishers’ systems◦ The ISBN used to identify both a Work and a Product◦ Not a huge problem in the world of physical products…

b i i h ld f di i l◦ …but a growing one in the world of digital ones…◦ …where we can add the problem of using the same identifier

to identify two (or more) different products G t l it b t ti ll dd t th h ll Greater granularity substantially adds to the challenge◦ Granularity of digital use….many different products (different

ebook formats, different channels, different devices)G l it f di it l t t diff t it f t t◦ Granularity of digital content…many different items of content used in different contexts (eg the same content used in many different learning objects)

Beginning at the beginning...

Atoms/bits “I made it”ManifestationFixation

Perceivable

Abstracted to

Expression Spatio-temporal

Actions “I did it”Fixed in

Expression temporal

Expressed inAbstracted to

Conceptual

Th ht “I i d it”

Abstractionaka “Work”Thoughts “I conceived it”

Digital Asset Management as an example of the requirement for work identification in the publishing process.

ManifestationsAbstractions ItemsWorksClasses of perceivable assets Individual perceivable assetsAbstract assets or “content”

Most recent standard metadata and identifier systems (eg ONIX, FRBR, DOI, RDA, DDEX, ISTC) recognize some variation of this model, though the exact terminology may vary

Fixed, made of atoms

Content

Physical Physical

Fixed, made of bits

Abstract, distinct from any specific perceivable manifestation

Content

Digital Digital

Manifestations are classes of physical or digital assets with common identifiers and identical attributes. For example: the class of all books with the same ISBN is viewed as a single Manifestation. An individual copy of the book is an ItemAn individual copy of the book is an Item.

ManifestationsAbstractions ItemsClasses of perceivable assets Individual perceivable assetsAbstract assets or “content”

egthe words of a book or articlethe image in a photograph

the figures and layout of a tablethe figures and layout of a tablea logo

the design of a chaira graph

eg, a class ofprinted book

music CD

eg, an individual copy of a printed book

music CDContent

Physical Physical

or the complete contents of a volume of a journal

a magazine

music CDleaflet

sculpture

music CDleaflet

sculpture

Content

Digital Digitalg

a book of photographs and maps a series of books

a websitean academic course pack

a tv soap opera

eg, a class of.pdf of an article

.mp3 of a sound recording.exe of a program

.gif of a photograph

eg, an individual copy of.pdf of an article

.mp3 of a sound recording.exe of a program

.gif of a photograph

This simplified example shows just one asset and relationship of each type. In practice there may be multiple assets of all types with relationships at all levels including between assets of the same type

ManifestationsAbstractions Items

levels, including between assets of the same type…

Classes of perceivable assets Individual perceivable assetsAbstract assets or “content”

Physical Physicala copy of

Hardback printedition

Content of book

Content Digital Digitaldownload

This is also simplified, but starts to show the complexity of identification and relationship that e ists

ManifestationsAbstractions Items

relationship that exists...

Classes of perceivable assets Individual perceivable assetsAbstract assets or “content”

Physical

Print on Demand

Content Digital Physical

Content

Photograph TIFF

Digital

PDF

Customer copy

DigitalDigital

Content of book

Content

Ch

Content

F d

Digital

Digital

XML

DigitalDigital

Individual files in DAM

System

Physical Physical

Chapter Foreword

Content

Adobe Ebook

MS Ebook

Digital

Hardback printedition Archive copyIllustration

MS Ebook

...identifying creations.

ManifestationsAbstractions ItemsClasses of perceivable assets Individual perceivable assetsAbstract assets or “content”

Physical PhysicalLibrary shelf mark

ContentISTC

ISBNmark

Individual copyHardback printedition

DOIGS1 SGTIN (RFID)

Digital Digital

ISBN

Filename

Content of book

ebook File onISBN File on individual computer

DOI (any)

Some standard identifierssuitable for identifying assets of different types

ManifestationsAbstractions ItemsClasses of perceivable assets Individual perceivable assetsAbstract assets or “content”

Physical Physical

Content

ISBN (books)ISMN (sheet music)UPC (products)EAN13 (products)DOI (any)

Individual copy

DOIGS1 SGTIN (RFID)

Digital DigitalISTC (words)DOI (any)PII (articles)ISSN ( i l )

DOI (any)

File onISSN (serials)ISAN (audiovisual)ISWC (music)

DOI (any)ISBN (products)

File on individual computer

DOI (any)

When there is a need to communicate across organizational boundaries – within a supply chainorganizational boundaries – within a supply chain

Internal requirements for interoperability between systems are not the same as external communication requirementsrequirements

Never allow your operational flexibility to be limited by the limitations or requirements of existing standards◦ You must be able to identify what you need to be able toYou must be able to identify what you need to be able to

identify within your own systems how, when and where you need to identify it!

◦ …and to be able to identify the same entity with standard id tifi h thi i i tidentifier when this is an appropriate

Prefer standards for external communications◦ Minimise confusion and complexity in the supply chain

...the solution to textual work identification?

The purpose of the International Standard Text Code (ISTC) is to enable the efficient identification of textual works. The ISTCenable the efficient identification of textual works. The ISTC provides a means of uniquely and persistently identifying textual works in information systems and of facilitating the exchange of information about those works between authors, agents, publishers retailers libraries rights administrators and otherpublishers, retailers, libraries, rights administrators and other interested parties, on an international level.

The ISTC may be applied to any textual work, whenever there is an intention to produce such a work in the form of one or more pmanifestations. It provides an identification data element for applications that record and exchange information about textual works and related manifestations. For example, the ISTC may be used for the purposes of collocating subsequent manifestationsused for the purposes of collocating subsequent manifestations of the same work or derivations of the same work in applications involving electronic rights administration or information retrieval.

Textual works that are eligible for an ISTC include any di ti t b t t tit d i tl d fdistinct abstract entity, predominantly composed of a combination of words, that can be described to satisfy the ISTC metadata requirements. In order to be assigned an ISTC, the declared metadata for any textual work at the time of registration shall contain at least one element pertaining to the work itself that p gdistinguishes it from every other textual work to which an ISTC has already been assigned.

If two entities share identical ISTC metadata they If two entities share identical ISTC metadata, they shall be treated as the same textual work and shall have the same ISTC.

Different users may have different functional requirements – so may have a different view of therequirements so may have a different view of the necessary granularity:◦ Publishers may need to collocate all the different

manifestations of the same edition of a book they ypublish◦ Retailers may need to do the same…or may need to

collocate all the different editions of “the same work” from many different publishers◦ Librarians may need to collocate the same work, but

distinguish between what FRBR calls different “expressions” of “the same work” ◦ Rights management organisations may need to

distinguish between different versions of “the same k” b f diff i i h hiwork” because of differences in rights ownership

Yes… but at the expense of rather more hi ti t d t d t t thsophisticated metadata management than we

are used to◦ Many relationships have to be created and managedy p g

And what about the problems of “fragments” of text…◦ or photographs◦ …or photographs

Serious system implications for everyone who needs to manage the ISTC

ll d f d bl ISTC will need to find a sustainable economic model

...why have we run into problems & how can we escape?escape?

ISBN system devised in late 1960sI iti ll i l t d i th UK th 9 di it SBN◦ Initially implemented in the UK as the 9-digit SBN

ISO ISBN standard (ISO 2108) first published in 1970◦ UPC introduced in 1973, EAN-13 in 1977

Universally adopted as the key identifier for books in the supply chain (agencies in 170 pp y ( gcountries)

4th Edition of standard published May 2005◦ 13-digit ISBN – 1 January 2007◦ 13-digit ISBN – 1 January 2007◦ Explicit guidelines for e-books◦ Assignment to chapters/fragments

“A separate ISBN shall be assigned to each separate hi bli ti t diti fmonographic publication, or separate edition of a

monographic publication issued by a publisher. A separate ISBN shall be assigned to each different language edition of a monographic publication.”

“Different product forms (e.g. hardcover, paperback, Braille audio-book video online electronicBraille, audio book, video, online electronic publication) shall be assigned separate ISBNs. Each different format of an electronic publication (e.g. “ lit” “ pdf” “ html” “ pdb”) that is published and.lit , .pdf , .html , .pdb ) that is published and made separately available shall be given a separate ISBN.”

Ease of trading◦ Most book trade e-commerce systems require ISBNs◦ Certainty of identification is critical for effective e-

commerce Ease of discovery of the different formats

available ◦ Bibliographic databases require ISBNs and users do not◦ Bibliographic databases require ISBNs and users do not

want to be tied to one channel Collecting detailed sales/usage data◦ If separate formats are not identified in a standard way,

sales and usage data by format cannot be easily collected

“We only “publish” one generic format (e.g. epub) and assign an ISBN to that”.epub) and assign an ISBN to that

“We are not responsible for formats provided by third part intermediaries”

“We don’t care whether or not different product formats are listed in bibliographic databases.”

“Our hardware-led channels do not require Our hardware led channels do not require standard identifiers and customers will find our books through their preferred platform.”“Our system requires us to manually create and Our system requires us to manually create and manage separate ONIX records for each ISBN we assign.”

It avoids an explosion of identifiersThink of all the numbers you might need when you◦ Think of all the numbers you might need when you multiply the different potential permutations of content by the number of different formats

But will it work in the supply chain? Not everyone But will it work in the supply chain? Not everyone thinks so“Each e-book title should have a unique ISBN for its format and for its endor This is necessar toformat and for its vendor. This is necessary to allow librarians to easily discover who is supplying e-books, in what format they are available and th h hi h d th i th ”through which vendors they can acquire them.”

JISC Collections(UK) Consortium for Common Information Infrastructure (the Netherlands)

1 Use proprietary product identifiers in the1. Use proprietary product identifiers in the

channel

2. Have someone else apply ISBNs in the

channel

3. Introduce yet another new identifier…like the music industry hasy

AdvantagesSome vendors already apply proprietary identifiers at the◦ Some vendors already apply proprietary identifiers at the level of individual SKUs, so no additional work

◦ Publishers don’t need to bother with proliferation of new identifiers, and can simply issue an “ebook ISBN” (againstidentifiers, and can simply issue an ebook ISBN (against which vendors report)

Disadvantages◦ Further along the chain (eg in libraries) the identifiers will u t e a o g t e c a (eg b a es) t e de t e s

have no meaning (and may be impossible to manage)◦ The information available to publishers collecting data

simply against a single ISBN may be inadequate The worst of all possible worlds?◦ Identifiers which look like ISBNs but are not A systems driven solution – semantic and technical chaos

Advantages◦ Identifier familiar throughout the chain◦ Identifier familiar throughout the chain◦ Publishers don’t need to bother with proliferation of new

identifiers, and can simply issue an “ebook ISBN” (against which vendors report)

Disadvantages◦ Potentially, considerable confusion – is the channel the correct

point of granularity?P bli h d l di lik h id f l b i◦ Publishers deeply dislike the idea of someone else being allowed to identify “their books”

Nevertheless, some wholesalers are now moving towards getting an ISBN prefix and assigning their owntowards getting an ISBN prefix and assigning their own ◦ “We will always prefer the publisher’s format specific ISBN and

also will link the parent ISBN [???] to our own prefixed number”

The Global Release Identifier [GRid]Id ifi “b dl f Di i l◦ Identifies… “bundles of one or more Digital Resources compiled for the purpose of electronic distribution. It is not used to identify any specific P d t hi h t i h R l i di id lProduct which contains such a Release, or individual instances of the Release.”

Purpose: to manage the proliferation of products and the lack of an appropriate standard product identifier◦ The music industry has never had its own standardThe music industry has never had its own standard

product identifier◦ Has used UPC/EAN

SongISWC

RecordingRecordingRecording RecordingISRC

RecordingRecordingRecording Recording

Many recordings of the same song

SongISWC

SongSong

RecordingISRC

RecordingRecordingOther Recording

GRid

RecordingRecordingcontent

ReleaseGRid

ProductProductProduct ProductISRC!Prop. UPC

Many products with the same content, but different technical characteristics, permissions etc

Advantages◦ Clarity of identification

Disadvantages◦ Implementation costs and comprehension problemsImplementation costs and comprehension problems

The reality in the music industry◦ GRid adoption has been slow

ff l b l l d ff◦ Different labels are applying in different ways The labels have never been very disciplined in applying ISRC

◦ There is a huge amount of metadata required in reporting Not simply “10 copies of GRid 123 sold this month” Standard product identifiers also required?

Millions of books without ISBNs Millions of books without ISBNs◦ Should they be applied retrospectively?◦ By whom?Di iti ti f th illi b k ( ith d Digitisations of those millions books (with and without ISBNs)◦ Should they be given (different) ISBNs? ◦ By whom?

Critical decisions to be made by ISBN community as well as by those undertaking digitisationswell as by those undertaking digitisations◦ But note these two – completely different – potential

applications of ISBN should not be conflated

...the next one to watch?

From a draft of the “Draft International Standard” (DIS)( ) This International Standard specifies the International Standard

name identifier (ISNI) for the identification of public identities of parties; that is, the identities used publicly by parties involved throughout the media content industries in the creation, production, management, and content distribution chains.

The ISNI system uniquely identifies public identities across lti l fi ld f ti ti it d id t l fmultiple fields of creative activity and provides a tool for

disambiguating public identities that might otherwise be confused.

The ISNI is not intended to provide direct access to The ISNI is not intended to provide direct access to comprehensive information about a public identity but can provide links to other systems where such information is held.

Library name authority projects [VIAF] Rights management ◦ Across the media

Has implications (eg) for the Book Rights Registry◦ Has implications (eg) for the Book Rights Registry Requires the development of unique identities for

publishers and imprints Has potential retail application but not the

main driver“Other books by this author”◦ Other books by this author

Probably simply “watch this space”…◦ …and think about the implications for you from a

systems perspective if the system is widely implementedp

If you want to influence, you need to engage through your national body (NISO in the US)◦ Moving towards the end of the process

Some conclusions...

Identifiers support interoperability between systems; if all those systems are within the same organisation, then identifiers can y g ,(and should) be proprietary ◦ Nobody (other than you!) cares much about how your DAM

communicates with your distribution system or your royalty system Don’t allow external constraints to dictate internal system

requirements◦ Standards are for external not internal interoperability

Short term system constraints are a poor basis for determining a strategygy◦ However persuasive those arguments are at a time of financial

constraint You can identify things with the same identifier as long as you

are happy always to treat them as being “the same thing” – but h d d h b h hwhen you need to distinguish between them, using the same

identifier will cause you problems Whatever you decide will be extremely difficult to undo◦ Lumping is easier than splitting “after the event”…p g p g

As you need to communicate with other people’s systems, standard identifiers become increasingly helpful – particularlystandard identifiers become increasingly helpful particularly in supporting unambiguous many to many communication.◦ Common syntax◦ Common “identity model” What is being identified What are the granularity rules

No single constituency in the supply chain makes decisions about the identifiers that are going to be applied it requiresabout the identifiers that are going to be applied, it requires consensus◦ It is only supply chain pressure that will be effective in “enforcing”

the consensus…◦ …and that this may not be easy if not everyone wants standards

(for their own reasons)…◦ …and remember that some constituencies will be more powerful

than others

ISTC has to demonstrate that is has sufficient value to overcome inertia (and particularly the cost ofovercome inertia (and particularly the cost of appropriate metadata management) ◦ …and to demonstrate that it can properly fulfil the different

requirements of its very different constituenciesq y◦ Book Rights Registry may be the tipping point Work collocation and disambiguation lie at the heart of rights

management ISBN has to resolve some significant challenges if it is ISBN has to resolve some significant challenges if it is

continue to be an effective identifier for the next 40 years◦ It certainly isn’t dead but we could kill it (and will almost◦ It certainly isn t dead…but we could kill it (and will almost

certainly regret it if we do) ISNI, like ISTC, will need to demonstrate value◦ …but don’t forget that the value to others may be considerableg y

Do we need other identifiers?◦ Gain consensus around business requirements◦ Gain consensus around technical requirements◦ Then and only then specify the solution◦ Then, and only then, specify the solution

Don’t take an existing identifier and try to use it for something for which it was not gdesigned◦ It won’t work

Y i k ti l t h◦ You risk creating complete chaos….

A quick audience poll...

Now that you’ve taken part in this BISGNow that you ve taken part in this BISG Webcast, what do you believe is the biggest

barrier to assigning ISBNs to digital products?barrier to assigning ISBNs to digital products?

Now that you’ve taken part in this BISG Webcast,what do you believe is the biggest barrier to assigningwhat do you believe is the biggest barrier to assigning

ISBNs to digital products?35.00%

15.00%

20.00%

25.00%

30.00%

0.00%

5.00%

10.00%

There are no Price of Perceived Information Current Current OtherThere are no barriers11.3%

(16 votes)

Price of ISBNs8.5%

(12 votes)

Perceived value (or

lack thereof) for my

business6.3%

Information / metadata

"bloat"33.1%

(47 votes)

Current workflows

make it difficult to

assign them19.0%

Current digital

business model(s)

don't necessarily

Other5.6%

(8 votes)

Results gathered during a live

participant poll.

6.3%(9 votes)

19.0%(27 votes)

necessarily require them

16.2%(23 votes)

Series1 11.30% 8.50% 6.30% 33.10% 19.00% 16.20% 5.60%

Mark Bide: [email protected]: www.editeur.org

Andy Weissberg: andy weissberg@bowker comAndy Weissberg: [email protected]: www.myidentifiers.com

Angela Bole: [email protected] b i bi

61

Website: www.bisg.org