Generalizing the International Children's Digital Library

poster / demo / art installation
Authorship
  1. 1. Benjamin B. Bederson

    University of Maryland, College Park

  2. 2. Patrick Rutledge

    University of Maryland, College Park

  3. 3. Alex Quinn

    University of Maryland, College Park

Work text
This plain text was ingested for the purpose of full-text search, not to preserve original formatting or readability. For the most complete copy, refer to the original conference program.

The International Children’s Digital Library is a
website of exemplary free children’s books from
around the world (www.childrenslibrary.org). Since
we launched the website in 2002, we have recently embarked
on an effort to generalize what we have learned,
creating a version (to be made open source when it is
finalized) that is suitable for more general content and
a more traditional adult audience. We also have been
exploring mobile deployment – providing access to the
ICDL’s picture books with an interface that displays
readable text in context on Apple’s iPhone. Together,
these efforts demonstrate how we can learn from one
successful project to expand into other areas of content
and platform, and in this paper we attempt to summarize
the core lessons learned that can be applied elsewhere.
Introduction
The International Children’s Digital Library (ICDL) is
an established electronic archive, providing children and
their adults easy access to thousands of children’s books
in almost 50 languages which can be read online for free.
Over the six years since we deployed the ICDL, we have
learned a lot about what it takes to make a usable digital
library (Bederson 2008). Motivated by a need of the
Boston Public Library and the Knowledge Commons
(formerly the Open Content Alliance) to offer a highly
usable range of interfaces for the millions of books they
are scanning, and by the recognition that people are accessing
more and more books on mobile devices, we decided
to start branching out, and applying those lessons
to other content domains, audiences, and platforms.
Content
The books in the ICDL are all digitized versions of traditionally
printed paper books which are selected for children
ages 3-13. Almost half are picture books. Due to the
narrow range of content, we have been able to manually
catalog the books in the library with a fixed ontology of
just under 300 categories in a hierarchical structure (i.e.,
Appearance->Format->Picture Books). We use a combination
of traditional categories (such as “Short Story”)
and more customized categories that we developed in the
course of working with children (such as “Book cover
color” and whether the books are “Happy” or “Sad”).
In order to design the ICDL interface to support a broader
set of content, we had to rethink how we categorized
the books and presented those categories. A few hundred
categories do not offer a fine enough distinction for
millions of books (Baker 1996). In fact, the standard
MARC records of the Library of Congress use many
thousands of categories. The traditional solution is remove
the focus on categories, and to make textual search
the primary mechanism for searching for books in large
collections. However, this gives up the power of categorization,
especially when the person has some particular
attributes they want to specify, or if, for example, a
teacher wants to specify a set of categories that might be
useful for their students to use for more focused searching.
Another scaling problem with our original design is
that we used manually created icons for each category.
This was a lot of work for a few hundred categories, but
impossible when there are many more.
We came up with a new solution (Figure 1) that lets the
user choose which categories are displayed in the primary
search window and integrates keyword search, the
search results, and a preview of the book. A tightly coupled
advanced search (not shown) adds several features,
including a more hierarchical display of categories,
counts of how many books are in each category, explicit
search fields for title & author, and more. Audience
Our aim of supporting adult users and not just children
also gave us some flexibility in designing this prototype.
Our research with children has led us to understand the
importance of using large objects to click on the screen,
fewer abstractions in the interface and a generally simpler
visual display (Hutchinson 2003). The interface we
created thus correspondingly uses more text, smaller hit
targets, and a visually denser display.
Platforms
In the same spirit of trying to make books available to
more people, we designed a version of ICDL for the Apple
iPhone platform. This platform is unusual because
while small, it has a relatively high resolution display
and powerful graphics that enable smooth animated transitions.
We built an application that gives offline access
to four picture books. It uses the “ClearText” technology
that ICDL developed for making text legible on small
displays (Quinn et al. 2008) and smooth animated transitions
that make it easier to understand where you are in
the interface and book as the reader navigates a complex
information space on a very small display.
Concluding Thoughts
As the trends of increasing technological capabilities
and broad increases in access continue to occur, we must
refine and innovate the way we deliver access to books.
It is crucial that people have an unfettered ability to find
what they are looking for, and to deeply engage with the
content of books once they have found it. The efforts described
here show two approaches to achieving this goal.
The lessons from the web interface include the importance
of building an all-in-one interface with varying
levels of search complexity along with search results
and book previews all integrated into a single screen. A
related lesson is that you can’t ignore issues of scale in
your design, and that one important approach is to couple
a very good and simple initial experience with the
possibility of significant end-user customization.
The lessons from the iPhone interface are that you must
re-think what you are offering for this tiny and mobile
package. You cannot and should not do everything that
you offer on your full website. It was a tough call for
us, but we completely eliminated search! The other issue
is that while difficult, you must balance engagement
and usability. We tried to use interesting animations to
keep the experience playful while not getting in the way
of reading – and we also spent a lot of time creating an
interface that supports the core mission of the application
– which is clear and legible text for easy reading. As
the application is just being released now, time will tell
whether we were successful with these goals.
References
Baker, S. (1996). “A Decade’s Worth of Research on
Browsing Fiction Collections.”
In Kenneth Shearer (Ed), Guiding the reader to the next book. New York, NY: Neal-Schuman Publishers, Inc.
45-72.
Bederson, B.B. (2008) “Experience the International
Children’s Digital Library”, Interactions Magazine,
ACM Press, 50-54.
Hutchinson, Hillary Brown (2003). Children’s Interface
Design for Hierarchical Search and Browse. In proceedings
of ACM SIGCAPH Computers and the Physically
Handicapped. 75: 11-12
Quinn, A., Hu, C., Arisaka, T., & Bederson, B.B.
(2008) Readability of Scanned Books in Digital Libraries,
in proceedings of ACM CHI (CHI 2008), ACM
Press, 705-714.

If this content appears in violation of your intellectual property rights, or you see errors or omissions, please reach out to Scott B. Weingart to discuss removing or amending the materials.

Conference Info

Complete

ADHO - 2009

Hosted at University of Maryland, College Park

College Park, Maryland, United States

June 20, 2009 - June 25, 2009

176 works by 303 authors indexed

Series: ADHO (4)

Organizers: ADHO

Tags
  • Keywords: None
  • Language: English
  • Topics: None