2. *
*The Archival Description Working Group is composed of
representatives from DPLA Partner institutions as well as
national-level experts in digital object description and
discovery.
The group will explore solutions
to support both item-level and
aggregate-level approaches to
digital object description and
access. The group will be charged
with developing recommendations
for DPLA as well as developing
any data models or tools as
appropriate. The working group will
explore aspects of both object
description and access through
discussion, review of prior research,
and examination of exemplary
implementations at other
institutions.
Charge. The New York Public Library.
3. *
*Incorporating Finding Aids or other collection
records into DPLA
*Modeling or mapping EAD as DPLA SourceResource
objects
*Putting a stop to
either item-level or
aggregate-level
description practices
Stop the Madness!
University of Southern California Libraries.
4. *
*Recommendations for aggregated archival objects
*What contextual information is useful for items
*What structure (item- or aggregate-level) is useful in
what situations
*How should objects be
displayed in context in an
aggregation
*Sources for context of digital
objects
*EAD and archival notions of
collection
*Thematic and other notions of
collection
Easter Egg Race.
Nyack Library.
Empire State Digital Network.
5. *
Gretchen Gueguen (Chair/DPLA), Jodi-Allison Bunnel (Orbis Cascade Alliance),
Mark Custer (Yale University), Bradley Daigle (University of Virginia),
Jacqueline Dean (University of North Carolina, Chapel Hill), Max Eckard
(University of Michigan), Ben Goldman (Penn State University), Leigh Grinstead
(LYRASIS), Kris Keisling (University of Minnesota), Adrian Turner (California
Digital Library)
RodeoCommittee.UintahCounty(UT)Library.
MountainWestDigitalLibrary.
6. *
Shawn Averkamp (New York Public Library), Erin Hawkins (World Digital Library, Library of
Congress), Sheila McAlister (Digital Library of Georgia), Sandra McIntyre (Mountain West Digital
Library), Anne Van Camp (Smithsonian Institution)
HallettesvilleVFDCelebrationCommittee.
ThePortaltoTexasHistory.
7. *
*Research Phase (10/1/15 – 2/29/16)
*Literature Review
*Environmental Scan
*Summary/synthesis
*Design Phase (3/1/16 – 6/30/16)
*User Scenarios
*Metadata Analysis
*UI Analysis
*Writing Phase (7/1/16 – 7/31/16)
*Whitepaper
*Metadata Tools Blueprint of Victory.
National Archives and Records Administration.
8. *
* Literature Review
* Jodi Allison-Bunnell, Elizabeth Yakel, and Janet Hauck, “Researchers
at Work: Assessing Needs for Content and Presentation of Archival
Materials” Journal of Archival Organization. 9:2 (Fall 2011): 67-104.
* Bardi, et.al. Recommendations for the representation of
hierarchical objects in Europeana. Version 1.0. 2014.
* Wickett, et.al. Modeling Cultural Collections for Digital Aggregation
and Exchange Environments. Center for Informatics Research in
Science and Scholarship. 2013
* Wim van Dongen. A workshop on how to display hierarchical objects
and their metadata in Europeana. Europeana Metadata Workshop.
2010
* Oksana Zavalina. “Contextual Metadata in Digital Aggregations:
Application of Collection-Level Subject Metadata and Its Role in User
Interactions and Information Retrieval.” Journal of Library
Metadata. 11:3-4. (2011): 104-128.
* Environmental Scan
9. *
*Define what you mean when you say collection
*Collection information is useful when evaluating
individual digital objects
*Confusion can happen when
it isn’t clear that the item
is part of the collection,
or how it fits in
*Finding or mapping the data
can be difficult
Children's Dept., basement work room, showing Charles
K. Miles accessioning books.
Public Library of Fort Wayne and Allen County.
Indiana Memory.
14. *
Things to Build On:
*Collection context at brief and
full record
*Browsing by collections
*Mix of item- and aggregate-level
objects
Things to Improve On:
*Need to improve on transition
between finding aid contexts and
digital repository contexts.
A family outing along the river. Woman with
binoculars is likely Marie Donner.
Beaufort County Library.
South Carolina Digital Library.
15. *
*User Scenarios: http://bit.ly/userScenarios
*Brief “stories” that identify
*A type of user
*A motivation
*A desired interaction with data
*Additional work identifies
*Metadata needs to make the scenario possible
*UI needs to make the scenario possible
*Example implemenations
16. *
Juliet finds an item during a
geneaology search that confuses her.
The record contains information about
the collection to which it belongs,
which helps her make sense and
expand her search.
Desired Interaction:
*It should be apparent to users when
they find an item/s that these
materials are part of a larger
collection.
“Romeo and Juliet.”
Memphis Public Library.
Digital Library of Tennessee.
17. *
Stanley does a search related to his
personal interests. Collection
information in the brief record result
helps him choose which record to
explore further.
Desired Interaction:
*Users should know as soon as they
search that items are part of
collections and should be able to act
on that knowledge.
Stanley Ketchell.
The New York Public Library.
18. *
Melody is familiar with a collection
and wants only search results from it.
She narrows her search results by
collection membership.
Desired Interaction:
*Users should be able to refine and
limit their searches by membership
in collections.
“Melody in F” Sheet Music.
National Museum of American History,
Kenneth E. Behring Center.
19. *
Kurt searches and finds a record for a
folder of letters. The collection
information in the record helps him
understand what it is and he explores
further.
Desired Interaction:
* Users should understand when objects
are described using a traditional
component-level archival-style
descriptions
Into the Library.
The University of Washington.
20. *
Augusta is an artist looking for visual
material. She finds lots of items with
similar descriptions from an archival
collection. The collection information
helps her make some sense of the
results, but since these are images, she
is pretty satisfied with her search
anyway.
Motivation
*Users should be presented with
appropriate metadata for objects, and
this level of metadata and context
may not be the same for all objects
and collections. Augusta Savage, sculpting.
National Archives and Records Administration.
21. *
Hiro finds an item that he is
interested in. The metadata implies
that it is not only part of a
collection, but part of several sub-
groupings. This helps him determine
that he wants to explore the
collection more fully.
Desired Interaction:
*Users may be presented with
information that helps them
makes sense of where the item
belongs within a collection if the
collection structure or arrangement
is meaningful.
Unidentified Japanese Student.
Los Angeles Public Library.
California Digital Library.
22. *
Nadima finds an interesting item.
From the record she discovers that
it was used in an exhibit, which
also contains materials that
interest her.
Desired Interaction:
*Collection/context information
applies to different types of
collections including exhibitions
and primary source sets.
Muslim woman in class, South Asia or Middle East,
photograph by Jean Shifrin, 1992.
Georgia State University Libraries, Special Collections.
Digital Library of Georgia.
23. *
Marshawn wants to find
collections of materials to
help his students do research.
He browses through collection
descriptions.
Desired Interaction:
*Users can go to DPLA and find
a collection that interests
them without doing a item
search.
An African - American male professor teaches a Math
class at Elizabethtown Community College.
University of Kentucky.
Kentucky Digital Library.
24. *
Ling does a successful search.
While viewing an item of
interest, she is presented with
links to other items from the
same collection.
Desired Interaction:
*A user can find similar
materials related to a retrieved
item by their membership in
the same collection.
Othello – Costume Rendering.
University of Illinois at Urbana-Champaign.
25. *
*Metadata Analysis
*Collection information to gather
*Sources/mappings
*UI Analysis
*Collection information to display
*When and where
An alley and stairs outside a garment shop
shows cloth scraps falling out of a barrel.
Cornell University.
ARTstor.
27. *
*Recommendations only
*Comment by Community
*Exploration by DPLA of
feasibility
Astronaut Carrying Experiment Packages.
National Air and Space Museum.
28. *
Detective Comics No 108.
National Museum of American History,
Kenneth E. Behring Center.
29. *
“It’s a Bear,” the Mascot at Camp Greene. Charlotte, N.C.
The University of North Carolina at Chapel Hill.
North Carolina Digital Heritage Center.
30. User Scenario 1
* Items are part of
collections
User Scenario 2
* Collections are visible in
search results
User Scenario 3
* Searches can be limited
and refined by collection
membership
User Scenario 4
* Aggregate description is
enhanced by collection
information
User Scenario 5
* Items with little to no
descriptive info work in some
contexts, but are always
improved with collection
information.
User Scenario 6
* Materials that are part of a
hierarchy can contain that
information.
User Scenario 7
* “Collections” can include
things like exhibits.
User Scenario 8
* Collections should be browse-
able
User Scenario 9
* Recommender algorithms
should include collection
membership
Notes de l'éditeur
Welcome to the Archival Description working group update.
As you may remember, back in September DPLA announced that it was forming this group and asked for volunteers.
Our charge was to explore solutions to support both item-level and aggregate-level approaches to digital object description and access. The group was tasked with developing recommendations, in the form of a whitepaper, as well as any data models or tools as appropriate.
The methods the group were to employ included review of research and examination of existing implementations.
http://dp.la/item/ba7002ea1ef5b46385dea07bb45af170
This caused some confusion initially and before really beginning, we stopped to define for ourselves what we weren’t going to do, which was
To talk about incorporating entire finding aids into DPLA
To create data models or mapping for EAD into DPLA sourceResource objects
We also aren’t going to suggest that all digitized objects be described at either an item-level or aggregate-level, therefore trying to put a stop to one or the other practice
http://dp.la/item/498035cbe2aac92dab8d0f5757d54dd7
Instead, what we are going to do is to create recommendations for access to aggregated archival objects in the DPLA context
We are going to talk about what contextual information is useful for item-level records
What kinds of data structures are useful in what situations
As well as how objects can be better displayed in context in aggregations like DPLA
We will be looking at both the traditional archival notion of collection
As well as the more recent thematic, or „digital „ collection notion of collection.
In short, we are trying to balance the needs of collection of item
http://dp.la/item/c21f0e9a59f86dd4b6850097b8d9cd07
With that in hand, the group formed, the members are
Myself
Jodi Allison Bunnel
Mark Custer
Bradley Daigle
Jacqueline Dean
Max Eckard
Ben Goldman
Leigh Grinstead
Kris Keisling
And Adrian Turner
http://dp.la/item/652871697c760991cc2efc4bb3a3374f
In addition, we named an advisory board. These members will help with reviewing our recommendations and providing an initial review. The board includes
Shawn Averkamp
Erin Hawkins
Sheila McAlister
Sandra McIntyre
And Anne Van Camp
http://dp.la/item/212a7e8bfd7d066760c6e4c3b39f242f
With the membership in place we set about creating a work plan for ourselves.
The first phase would involve research, both of the literature and of actual projects. We completed this phase at the close of February.
At the moment, we are in the middle of the Design Phase, and I will talk more about that today.
After the design phase is complete, we will enter a writing phase later this summer. These date are a bit squiggly, I assure you, but we are hoping to have something in draft form at least by the time of SAA which is in early August this year.
http://dp.la/item/be0cf9a26348a716afb36781b6342cc0
Our research phase started with a Literature Review. There wasn’t much out there that was specifically on this topic actually, but we did find some very useful reports out of Europeana, Archives Portal Europe, and the IMLS DCC project. We will be including a literature review in the whitepaper on what we found from these sources, but the general overview was
First of all, that collection means a lot of different things in a lot of different contexts, so define what you mean before you start working on them.
Generally, collection information is very useful to users when evaluating individual digital objects
However, confusion can happen when it isn’t how the item fits into the colleciton or that it is even related to the collection.
Finally, finding or mapping that collection data to the item can be difficult.
Those were basically the highlights.
We also spent time looking at a lot of different digital projects that used collection data in different ways.
http://dp.la/item/2a70a6d4d6accde09ae12e335f3b44a4
Some of them were more traditional find aid projects that incorporated large numbers of digital objects like the UNC Chapel Hill’s Southern Historical Collection, or the Archives of American Art’s collections online.
We obviously liked the way that these sites offered a contextual view of materials, but we found the intersection between these items and the rest of the digital library was restricted. These items either didn’t get into the digital library at all, or did with very minimal metadata.
We also looked at more “traditional” digital collection style sites that incorporated collection- information including the University of Minnesota’s Umbra site as well as Calisphere.
We found a lot of features to like in these two projects in particular, including the way that collection information was included in browseable search facets, integrated into item-level records, and used to suggest further items of interest.
We managed to find a few sites that were kind of like hybrids and had archival components likes series, boxes, and folders alongside other digital objects like image and films. Two were Yale University’s digital collections which have incorporated some entire collections, as well as the Rock and Roll Hall of Fame’s Library and Archives Catalog.
Each of these were interesting in the way they represented objects that were either folders (or aggregations to use a generic term) or items. The decision of whether to use which type of description (aggregate or item) was based on the content and what worked for it. These types also included features we really like such as being able to facet by collection entity, and seeing collection information as part of the item record and brief results.
Finally, we also looked at other existing aggregations to see how they were incorporating collection information. The Archives West portal is an aggregation of finding aids, but incorporates digital items into displays. Europeana has implemented collection membership in it’s item, here I’ve search for a collection object, but contextual information is scant.
These two aggregators in particular were very different, one being focused on finding aids and the other on objects. They do model, however, ways to handle the distinction between membership in the “collection” belonging to an institution, and membership in some other type of collection.
With all of this data gathered, we moved on the phase we are currently in, Design.
So at the end of our research into existing projects we found both things to build on like implementing collection context at brief and full record views, browsing by collections, and ways to handle the mix of item-and aggregate-level objects.
We also found things to improve on, particularly the transition between finding aid contexts and digital repository contexts
The first part of our design work has centered on defining User Scenarios.
These are brief stories basically that identify a type of user, a motivation, and a desired interaction with data.
These scenarios help us define what metadata and UI changes would be needed in order to create these implementations. Each of the scenarios is based on something that we saw in either the literature or the other sites we looked at, so the final draft of these scenarios will include example implmentations wherever possible.
For the rest of this session, I’d like to walk you through our scenarios and at the end maybe get some feedback on whether or not we’ve missed any big areas.
Scenario 1
Juliet finds an item during a geneaology search that confuses her. The record contains information about the collection to which it belongs, which helps her make sense and expand her search.
Desired Interaction:
It should be apparent to users when they find an item/s that these materials are part of a larger collection.
http://dp.la/item/424028abb5f3157579b3c6c364e62a9a
…in other words, if metadata is going to be done at an item level with very little unique metadata, the collection information should be included for context. Furthermore, the institution should consider when this level of description is adequate (in this case with images, it worked) and it when it isn’t (perhaps not great for less visually distinct materials)
We know that this data exists…
So maybe we can make it better…
And maybe we can provide some recommendations for when to use it and when not to in the future…
http://dp.la/item/d482ae584c75334eb1ceb8419ad8e2c6
So those scenarios set up a kind of framework that we can begin to use to analyze metadata and interface. From these scenarios and our previous research we hope to determine what collection information we recommend gathering, as well as some analysis of how that data might be created and shared, and where it might live within the DPLA metadata environment.
In addition, we will be analyzing the user interface to determine what collection information should be publicly displayed, as well as where and when that should happen.
http://dp.la/item/d4337aac2f5ac886be6a0cbed8611c6c
Once that design work is complete, as I mentioned, we will release a whitepaper which will include those scenarios and recommendations, as well as any tools we develop like data models, crosswalks or wireframes.
http://dp.la/item/25a52ed49463716653b8a0eea15abf12
After the whitepaper is released there really aren’t any formal steps defined. The whitepaper is a recommendation only. There will be a period of public comment, after which DPLA will take some time to review the recommendations and explore whether and which are feasible for us.
http://dp.la/item/d6ce93a22f8b8c1ddf3c141d8e4e9424
So with the time we have left I’ll ask you all to be detectives and tell us what you think we’ve missed along the way…
http://dp.la/item/774395f4a1618fd442f67f0331cdc77e