MAM? DAM!

DAM: Digital Asset Management
MAM: Media Asset Management
CMS: Collection Management System
aka: “Our Archival Catalogue”

Disclaimer: DAMs or MAMs may be used for preservation, but are sometimes/often not designed or intended for use in a preservation context, but merely to store and handle “digital assets”: from regular office files (documents, images, etc) to managing in-house assets of larger companies. And some of these systems were then “also” used by archives.

The term “Collection Management System” usually indicates that it was more likely intended to be used in a preservation context, such as museums for example - where it may be used beyond digital: To handle physical collections even, like books, chairs, or anything. And files ;)

You may see the term “Collection Management System” being used interchangeably with DAM or MAM by the preservation community.

(Note: In the IT world, the abbreviation “CMS” is more commonly known as: “Content Management System” which is something completely different. In order to save some screen space, I will use the abbreviation “CMS” in these slides however instead of typing “Collection Management System”)

A DAM is usually the generic version of MAM - and sometimes the borders between “is it a MAM? is it a DAM?” are fuzzy and unclear, because they are so closely related.

Typical for “classic” DAMs: Often trimmed and designed for handling “2D material” (documents, images). When it comes to audiovisual, they’re mostly inadequate or not suitable for archive-suitable quality media handling.

MAMs are usually better suited for handling media. For example:

auto-generating access/preview copies suitable for low-bandwidth/internet/browser access.
image area annotation
time based annotation (“markers”)

But watch out! Even systems designed to handle AV media, are often not taking too much care about preservation/archival aspects. It’s not uncommon that “looks/sounds good enough!” is exactly where you’re at.

So please:

Try before you buy!
Don’t trust sales. Ever.

A short list

of some popular OpenSource CMS:

A loooong list…

of not only OpenSource CMS:

https://bits.ashleyblewer.com/blog/2017/08/09/collection-management-system-collection/

Criteria Overview

Controlled vocabulary lists
Import/Export abilities
Vendor/license dependency
Metadata Standards compliance
“Classic” vs “Entity relationships”
Does it do Web?
Access control
Failsafe options
Does it scale?

Your Wishes and Experiences?

Feature Wishlist?
Bad experiences (=avoid in the future)?
Good experiences (=want this again)?
Questions you’d like answered?

Controlled Vocabulary

“Controlled vocabulary schemes mandate the use of predefined, authorised terms that have been preselected by the designers of the schemes, in contrast to natural language vocabularies, which have no such restriction.”

Source: “Controlled Vocabulary” (Wikipedia)"

CoVoc Examples

Same-same 😇, but different… 😭

35mm = 35 mm = 35 millimètre
dup pos = duplicate positive
de = deu = german = German = alemán
yuv422p10le = YUV, 4:2:2, 10 bpc
Director = Directed by

Handling CoVocs

Choice / definition of terms?
Standard terms (e.g. ISO mapping) for: Language, Country, Region, Tech-Terms, etc.
Maintaining / augmenting terms?
Import / Export lists?
Identifier vs Label

Identifier & Label

Can the MAM store & handle this properly?

Label (for humans)	Identifier(for machines)
Albanian	`sq`, `sqi`, `alb`, …
German	`de`, `deu`, `ger`, …
Undetermined	`und`

Examples: ISO 639 (Languages), ISO 3166 (Countries),…

Shared CoVocs

Same vocabulary terms/lists can be used across different systems/domains.
Use existing CoVocs where possible.
Share/publish yours.
Be in contact/exchange with others.
Be ready for embracing trade-offs.
Can your MAM handle “dynamic” CoVocs?

Import / Export

Lists / CoVocs?
Cataloguing data? (DB)
Your files? (storage)
Value/term/schema mapping options?
Programming language & skills?

Access & Exchange: Data Formats

Linked (Open) Data

“linked data (often capitalized as Linked Data) is structured data which is interlinked with other data so it becomes more useful through semantic queries”

Source: Linked data (Wikipedia)

LoD is still a bit “new”, and there are many many many things to be considered, tested, found out, implemented, improved, etc. Yet, having isolated catalogues and maintaining all entitites individually, does not seem like a long-term sustainable approach.

However, many archives are still using conventional “flat” data models, local CoVocs - and are slowly transitioning towards shared CoVocs and Entity-Relationship models.

If you are interested in more hands-on practical experience with LoD and interoperability of data, you may want to get in touch with the FIAF cataloging commission, who is very actively charting the new territory of archives and LoD.

There’s a short summary from a FIAF LoD Workshop in 2019, which contains very valuable overviews and insights in a quick read :)

Linked (Open) Data

Wikidata

“[…] is a free and open knowledge base that can be read and edited by both humans and machines.”

Example: Earth (Q2) vs Earth (Q83697636)

MD-Standards: Some of Them

Dublin Core: Core fields (& their names)
METS: (MD container) descriptive, administrative, structural
PREMIS: A metadata framework
EBUCore: descriptive & technical (broadcast use case focus)
CEN EN 15907: Comprehensive description of cinematographic works
FRBR (“furbur”): Comprehensive description of bibliographic works
Mediainfo XML: Technical metadata (AV)

MAM Requirements for MD-Standards

Store ID/ISO values (or just label)?
Relationships between different schema?
Import/export schema support?
Field labels in UI vs “The Standard Terms”?
Watch out for “EEE”:
Embrace, Extend, and Extinguish

Data Structure Matters

“classical” catalogue = flat, table/cell thinking.

Better: Identify Entities and Attributes.
Then: Identify Relationships (and their “@”)
Entity Relationship Model (ERM)
Think in “Semantic Triples”:
Subject - Predicate - Object

Example elgrito.witness.org

Another Example: EN 15909

Great! Now what about their “film-related materials” (posters, documents, etc)?

Relationships

Forward: Agent1 is parent of Agent2
Reverse: Agent2 is child of Agent1

Relationships: @Attributes

(How well) can the system handle/search relationship-attributes?

Examples:

Which Work(s) feature which Agent(s) under the alias of …?
List all … related to a work, in order of their …?

Oh, btw: “Says who?”

Declare information sources?
At which level?
What if you have ≥ 1 sources?
Who edited which field when and why?
Could a source be “linked” (as relationship)?

Content annotations

Regular descriptive metadata
Image annotations?
Time-based annotations?
Import/Export of content annotations?
International Image Interoperability Framework (IIIF) support?

Reference between Catalogue and Files

How are catalogue entries related to files (and vice-versa)?
How (in)dependent of the catalogue can files be used/modified?

Files intact, but MAM gone?

Store identifying metadata

Failsafe options

Backup database.
Backup storage.
Setup MAM as Virtual Machine.
Test export options.
Documented export data format.
Documentation & Source Code.

Does it do Web?

Is the client browser-based?
How hard/easy/possible to serve a web frontend? (APIs, components, programming language, …)
Which operating system(s) are used/required?
Does it provide eg a “REST API”?

From Local Catalogue to Web Access

Which level of separated/interconnected do you want/need?

1 common database (DB)?
2 DBs, synchronized: how? + how often?
What about the content (files)?
Who may access (read/write/edit) what?

Access Control

Users, Roles, Rights, etc.

How important for your collection/setup?
How seamless? (SSO, LDAP, other?)
How secure?
How hard to maintain? (SysAdmin effort/needs/overhead/headaches?)

Does it scale?

Speed? (search, retrieve, data, files, etc)
Cache & tuning options?
Synchronize multiple instances?
Ongoing import/export? (in-house, external)
License conditions and impact?
Multi data schema in parallel?
Multi language support?
Migration to a “better scaling” system when needed?

Got Support?

Whom to call in case of …?
Option to switch/choose “your mechanic”?
Which conditions?
Support contracts?
Which license model?

Summary of Criteria / Considerations

CoVocs: Ability to have ID and label?
Multiple MD-schemas in parallel? e.g. Different entry types: film, image, documents, physical objects, etc
Can it be used to implement a certain Standard?
Import/Export abilities (data and files)?
Field mapping/translation?
Content (file) storage layout? (naming, folder structure, links)
Complexity of hardware/software setup? (For setup, maintenance and migration)
What if your MAM is gone/broken?
Artificial restrictions for future support/migrations?
Access levels: Who may access/edit/etc which parts?
Browser-based vs local client application
Got support?

In practice: Try it!

Get access to a demo installation.
Try to import/export lists/CoVocs.
Consider import/export with your data (layout).
Before you make a final choice.
Get cataloguers & technicians together.

The End

Questions? Comments?

Peter Bubestinger-Steindl

Peter @ ArkThis.com

Choosing and implementing a MAM system

MAM? DAM!

A short list

A loooong list…

Criteria Overview

Your Wishes and Experiences?

Controlled Vocabulary

CoVoc Examples

Handling CoVocs

Identifier & Label

Shared CoVocs

Import / Export

Access & Exchange: Data Formats

Linked (Open) Data

Linked (Open) Data

Wikidata

MD-Standards: Some of Them

MAM Requirements for MD-Standards

Data Structure Matters

Another Example: EN 15909

Relationships

Relationships: @Attributes

Oh, btw: “Says who?”

Content annotations

Reference between Catalogue and Files

Files intact, but MAM gone?

Store identifying metadata

Failsafe options

Does it do Web?

From Local Catalogue to Web Access

Access Control

Does it scale?

Got Support?

Summary of Criteria / Considerations

In practice: Try it!

The End

Questions? Comments?

Links