Designing an Untethered Educational Digital Library

Designing an Untethered Educational Digital Library

Andy Dong

University of Sydney

Wilkinson Building (G04) Sydney NSW 2008 Australia

+61 2 9351 4766 adong@

Alice Agogino University of California, Berkeley 5136 Etcheverry Hall

Berkeley, CA 94720-1740

+1 510 643 1819 aagogino@


Digital libraries, such as the SMETE Digital Library at UC Berkeley (, are quickly becoming mainstream tools for science, technology, engineering, and mathematics (STE&M) education at all levels. And while the vision exists for “anytime, anywhere” access to resources from educational digital libraries, the reality is that learners are tethered to these resources through connected computers in classrooms or homes. Because nearly 85% of students’ time is spent outside a formal classroom, transforming coincidental, daily events into meaningful learning opportunities would be expected to impact the level of science learning for children. This paper reports on a workshop held at UC Berkeley on the use of multimedia, wireless technologies and other information technologies for educational digital libraries and knowledge management. The paper also describes a prototype solution for an untethered digital library used to stimulate a discussion on nomadic inquiry and the potential for nomadic computing technologies to support the pursuit of personally-relevant questions and explanations linked to real world contexts and problems.


The Internet and Web are as commonplace in classroom education as chalk boards. As such, educational digital libraries, such as and the National STEME Digital Library are quickly becoming mainstream classroom tools. The availability of wireless networks and mobile handheld or pocket computers with wireless data access at price-points and form factors that make it possible for people to carry them everywhere challenges digital library developers to re-consider where and how learners will actually access the content. The assumption that the majority of users would access these digital libraries from classrooms or through broadband connections at home is, simply put, potentially misconstrued.

Two major technological trends point to this possibility. By the end of 2003, Strategis estimates that broadband wireless networks will serve 34 percent of all American households and 45 percent of all American business [12]. Direct Internet access to a handset was 100 million users in 2000 [13], and predicted to rise to more than 240 million data users by 2005 [3]. In East Asia, the penetration rates are staggering; for example, in South Korea, of the 50 percent with wireless data access, 60 percent have mobiles phones with data handling capability. There also exists evidence that US schools are showing a preference for handheld devices with wireless networking rather desktop or laptop personal computers with wired networking due to the lower capital and installation costs of the former.

Finally, it is well-established that informal (science) learning outside of the classroom is a significant factor in the development of scientific inquiry skills and motivation to learn [14]. Because nearly 85% of students’ time is spent outside a formal classroom [1], transforming coincidental, daily events into meaningful learning opportunities would be expected to impact the level of informal learning for children. Thus, combining the availability of emerging wireless access technologies and nomadic computing devices with educational digital libraries, opportunities for learning, tutoring, and collaborating outside of the classroom is potentially significant. However, the different technological characteristics of the various wireless devices and the myriad of wireless connection options only add to the challenging problem of effectively integrating the delivery of the digital library content to the wireless device. In this integration, the process of transforming the multimedia rich contents that reside at the digital library to fit the wireless client capabilities, available bandwidth and user preferences in a way that satisfies the informal learning environment is still an open problem.

However, taking learning materials authored and originally intended for use in classroom settings on “full-sized” computers into informal environments on wireless handheld devices (nomadic computing devices) raises numerous unanswered questions. Perhaps most fundamentally is whether the quality of services [7] of nomadic computing technologies and wireless networks affect user experiences leading to informal learning. Reduction in quality of the educational content may reduce (or possibly eliminate) any motivation for the end-user to learn. At the same time, desktop miniaturization is objectionable.

This paper presents a prototype service to deliver resources from an educational digital library to nomadic computing devices through wireless networks. The prototype system is enabling the evaluation of the support of wireless handheld computing devices and the role of STE&M education digital libraries on informal learning. Specifically, our research is examining how the introduction of digital libraries on mobile devices support conceptual change and whether digital library content in situ during information learning opportunities can lead learners to think about what they need to ask the “right” question(s). The prototype was presented during a workshop at UC Berkeley on the use of wireless mobile devices in educational digital libraries and knowledge management as a stimulus for discussion. The workshop drew a dedicated group of individuals from research, industry and foundations, including Dr. Alan Kay of HP Labs,



Randy Hinrichs of Microsoft Research, Heikki Huomo of Nokia and Milton Chen of The George Lucas Education Foundation.The workshop raised the principle that an ideal learning experience with an untethered educational digital library should include ways of immediate access to information, capturing and documenting an experience for later reflection, getting expert advice, and extending a physical place with a virtual overlay of information provided by the digital library.

The paper proceeds by describing potential contexts of use of the untethered digital library in “nomadic inquiry.” A prototype is described, including a discussion of the underlying technologies.A review of the prototype is then provided given the outcome of the discussions from the workshop.

2. An Untethered Digital Library for Informal Learning

While there are numerous ways in which wireless computing devices might be employed in an informal learning environment,we describe two example situations where an untethered digital library would be critical for supporting informal learning and associated research questions.

Scenario 1: “Everyday Learning”. Suppose children are on a school-sponsored field trip. They could access the digital library from the classrom, personalize their PDA for use with the digital library and select some educational material to enrich the trip.Suppose during a hike, they want to know how water quality affects the fish. To answer the question, the parent uses an Internet-enabled PDA to access the SMETE Digital Library and download a resource on the link between Trichodesium bacteria,nitrogen production and toxic algae. How do quality of service (QoS) [7] metrics such as refresh rate, color, sound, etc., affect the educational value of the resource? To what extent are digital learning resources that were intended for classroom settings appropriate for informal learning settings?

Figure 1 Children Engaged in Nomadic Inquiry

Scenario 2: Museum Learning. Children participate in an after school field trip to an interactive science learning museum,such as the Exploratorium in San Francisco, as shown in Figure 2,which already has a wireless network infrastructure [1]. To increase interaction with the exhibits, the museum installs a wireless network and gives the student, parent or guide wireless handheld computers so that they may ask questions about the exhibit to a “remote” expert, learn more about the scientific principles behind the exhibit, and record information about the

exhibits visited. How can content from the digital library be used to help children increase their scientific awareness? How would children find (e.g., end-user initiated search, “location aware”delivery of content directly from the digital library, etc.) content in the digital library related to the exhibit? Could they collaboratively find information or share knowledge with previous visitors?

Designing an Untethered Educational Digital Library

Figure 2 Children Using PDA at Interactive Science Museum These scenarios draw attention to the relationships among learning and associated learning environments within the context of creating meaning out of that environment from the knowledge base of the learner. Understanding the mediating role of mobile wireless devices in informal learning is a critical step to designing features into digital library learning environments that can facilitate the voluntary learning of a variety of cognitive skills such as inductive reasoning, critical inquiry and scientific argumentation.

Based on her work with the Electronic Guidebook project at the Exploratorium in San Francisco, Dr Sherry Hsi emphasized in her talk at the workshop that the role of the wireless handheld devices in using an educational digital library should be to support “nomadic inquiry, an inquiry-based approach to learning in which learners are moving both in physical space and across information landscapes.” [[9]]

3. A Prototype

There exists several functional approaches for delivering digital content intended for full-sized personal computers to nomadic computing devices with “thin pipe” network connections: “clipping”, transformation, and transcoding. More broadly, this problem of “content adaptation” is concerned with delivering content through multiple channels such as the Web,mobile phones, and personal digital assistants. The “null”solution, of course, is not to modify the content at all and deliver the content “as-is” to the nomadic device or perhaps to “cache and stream” the multimedia content [15]. Neither of these approaches appears palatable. Caching for disconnected operation is another possibility [10], but the possibility of learners knowing the material they’ll need a priori is unlikely. Ideally, all content could be adapted to an acceptable level of quality to be delivered through any channel. However, this is not feasible given the varying technological capabilities of the channels. As such,various solutions have been proposed and implemented.

“Clipping” techniques such as cached text data, rapid queries using SMS (short message service) and Blackberries [4], cell-phone based expert services, and “Web clipping” such

Designing an Untethered Educational Digital Library



AvantGO essentially remove most multimedia content (such as graphics and movies requiring plug-ins) and deliver essentially a text-only version of the content. Unfortunately, these technologies essentially force the learner into a didactic model of education of learning through one-way information delivery given their limited interactivity. A constructivist model requires much richer interaction with the digital library, one that’s not possible with text-only interaction modes. That is, more flexibility is required to suit the learning style of the individual and the pedagogical style of the learning resource, whether that style is didactic or constructivist.

Transformation methods employ style sheets appropriate to the capabilities of the remote device. The transformation may take place at the server or at the client. Regardless, this approach requires the remote device to inform the server of its capabilities such as screen size and color depth in the request message so that the server can select the appropriate style sheet to create display mark-up appropriate to the device. Typically, the device’s CC/PP (Composite Capabilities/Preferences Profiles) [8] in the HTTP header contain information relating to its technical characteristics (e.g., OS, CPU, memory, color, sound, screen size) and capabilities (e.g., MPEG decoder, application plug-ins). Due to the fairly high computational requirement on style sheet processing, with respect to mobile devices, the transformation occurs at the server. For desktop personal computers, this is not an issue; Microsoft’s Internet Explorer browser supports XML/XSL and XHTML/CSS processing. Web-based services use XSLT (eXtensible Stylesheet Language Transformations) as a language to express the transformations necessary for different devices and/or platform types, and then use the Web server to determine the proper transformer for a request and apply that transformation to XML, converting the XML into HTML or WML, for example.One common tool-kit is the Apache Cocoon [ ] an XML application server that abstracts data from presentation using XML to customize presentation for different clients. The drawback is that all content must be marked up in XML, and given that most educational content on the Web is already marked up in HTML, one must re-author them in XML or employ a commercial package such as JavaCC [ ] to convert and normalize HTML into XML. While these transformation techniques could handle multimedia content, developers must store multimedia in numerous sizes and formats for delivery as style sheets do not natively support multimedia content compression. Finally, these techniques utilize a one-size fits all similar approach. In other words, a mobile PDA using GPRS would receive the similarly formatted content as the same mobile PDA connected to an 802.11b network at a wireless “hotspot” despite the fact the latter has access to a much faster network.

Transcoding methods adapt the content to suit the requirements of the mobile device considering both the receiver’s device capabilities as well as network conditions without actually changing the content. As such, these techniques hold the best promise by which content may be dynamically modified based on the needs of the client. Also, the technology is ‘future-proof’ as it is adaptable to future technologies and delivery mechanisms.These transcoding techniques often operate as proxies which are intermediaries between the content server and the remote device.The transcoders perform content distillation and implement protocol extensions to adapt the content to the varying

computational capabilities and bandwidths of different client communication links. The TranSend [5] project at UC Berkeley,I B M ’s W e b I n t e r m e d i a r i e s p r o j e c t (W B I )[] and WebSphere T r a n s c o d i n g P u b l i s h e r [h t t p ://w w w] are examples of on-the-fly content adaptation using transcoding techniques.

As an example of the differences between these capabilities of these techniques, consider the following illustrations. To save on space, we leave it to the reader to visit from a full-sized personal computer connected to the Internet via a high-speed network connection and search for a learning resource. For example, search for resources on “disk drives.”

The “null” solution is not to transform any of the Web pages,which is shown in Figure 3 and Figure 6. This leads to a fairly undesireable result as the user is forced to scroll quite significantly in order to “see” the entire screen. Given limited short term memory, it would be difficult to remember where the search box is and the location of the links. A text streaming transformation produces similar pages but predominantly in text,as shown in Figure 4 and Figure 7. Since many multimedia learning resources contain multimedia elements, this might be a severe limitation on the usability of the resource. Figure 5 and Figure 8 show similar pages transcoded using a transcoding service named LiFRA provided by Prof. Aura Ganz of the Multimedia Networks Laboratory at the University of Massachusetts, Amherst. Notice that while all the elements of the Web page appear, the images have been reduced in size so that much of the page is visible in a single screen and the images have also been reduced in size and compressed to enable faster download.

The total downloaded size of the original home page is about 70.9kb, reduced to 32.1kb after transcoding. The size of the original find page is 43.9kb and 27.9kb after transcoding. The transcoding reduces the download time approximately 20%-25%while also adjusting the image and text size to the PDA’s screen

Designing an Untethered Educational Digital Library


Figure 3 Original SMETE Digital Library Home Page on

Pocket PC


Designing an Untethered Educational Digital Library

Designing an Untethered Educational Digital Library

Figure 4 SMETE Digital Library with WML Figure 5 Transcoded SMETE Digital Library Home Page on

Designing an Untethered Educational Digital Library

Pocket PC

Figure 6 Original SMETE Digital Library Find Page on

Designing an Untethered Educational Digital Library

Pocket PC

Designing an Untethered Educational Digital Library

Figure 7 SMETE Search Results in WML

Figure 8 Transcoded SMETE Digital Library Find Page on Pocket PC

4. Workshop Discussion

The workshop discussions centered on establishing design principles for handheld devices and learning content in the digital library to increase the engagement level with higher levels of interaction. The three dominant principles are reviewed here.

Dr Sherry Hsi proposed the following framework for organizing the types of learner interactions possible with a ubiquitous learning device.

1. Exploring the outdoor environment or artifacts in an environment

2. Getting more information (reference)

3. Documenting the environment (documentation)

4. Getting expert advice (recommender)

5. Carrying out an investigation (handheld as inquiry tool)

6. Collaborating with other participants (handheld as collaboration support tool)

7. Assessing learning (handheld as assessment support tool)Table 1 User Interactions in Informal Learning Contexts

Milton Chen reinforced items five and six in Table 1 as he expressed that the George Lucas Foundation Educational Foundation’s vision for education in the 21st century promotes student-centered learning with a focus on teamwork and collaboration. Thus, (Principle 1) knowing the type of interaction and to recognize the situation in which the learning takes place of the learner with the device should figure into the content adaptation as much as knowing the technical capabilities of the device.

Prof. Agogino and Prof. Mankoff noted though that the skill level of the student, both in terms of knowledge level and motor and sensors skills, could influence the accessibility of information. Prof. Mankoff emphasized that the mobile device should promote cognitive independence. For example, a student with low interest in animal biology might need a high quality video with audio whereas one who is already interested might be satisfied with an audio-only version or a clipped version with fewer frames. A blind student would also find the audio-only version suitable. As such, they proposed (Principle 2) that the digital library should perform content adaptation to take advantage of the learner’s profile, such as the in the meta-data contained in the IMS Learner Information Package (LIP).

Dr Alan Kay challenged the audience, though, to consider that the point of using computers in education, whether handheld devices or desktop computers, should be to enable children investigate and learn about deep ideas in ways that can not be accomplished in another medium. Dr Kay demonstrated Squeak, a media authoring tool which extends authoring in the dimensions supportable only by a computer, by showing how children learned about the spread of epidemics and linear systems furthered through relatively simple simulations. Randy Hinrichs and Mirjana Spasojevic commented that mobile devices were just the beginning of a trend of embedded computing in ordinary objects [11], and that interaction between the learner, the digital library, the handheld and ordinary objects should preserve and reinforce “everyday” learning experiences. As such, they offered that the operating system of mobile devices continues to restrict users to impoverished abilities for recording daily events which may lead to meaningful learning experiences. As a broader research goal and design principle and spirit of Kay’s Dynabook, they proposed (Principle 3) designing a mobile learning operating system that can facilitate the voluntary learning of a variety of cognitive skills such as inductive reasoning, critical inquiry and scientific argumentation.


The interactive nature of handheld devices and learning content in the digital library create the potential to increase engagement level with higher levels of interaction, which has shown to increase knowledge gain and thinking [1]. If interactive learning objects could be delivered on demand, a mobile handheld could be instrumental in supporting conceptual learning, inquiry-skill building, analytic experiences, or activities anytime, anywhere. This paper presented work-in-progress on an untethered educational digital library and design principles for its design based on a workshop on wireless and mobile computing in educational digital libraries and knowledge management.


The National Science Foundation partially funded this research work under DUE-0085878/DUE-0127580. Any opinions,findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation. The workshop at UC Berkeley was supported by a Digital Media Innnovations (DiMI) Grant and the Center for Information Technology Research in the Interest of Society (CITRIS). Prof. Rajit Gadh (UCLA) and Prof. Jen Mankoff (UC Berkeley) were co-PIs for the workshop with Prof. Agogino. Prof. Aura Ganz, of the Multimedia Networks Laboratory at the University of Massachusetts at Amherst provided a remote demo and the transcoded mockups shown in Figs. 5 and 8 A complete review of the workshop is available from


[1]Bannasch, S. (1999). The Electronic Curator: Using a

Handheld Computer at the Exploratorium. The Concord Consortium Newsletter, Fall. Retrieved from:


[2]Boisvert, D.L. & Slez, B.J. (1995, September). The

relationship between exhibit characteristics and learning-associated behaviors in a science museum discovery space.

Science Education, 79(5), 503-518.

[3]Cahners Business Information. (2000). Making money in M-

commerce: Partnering, platforms and portals [Electronic version]. Cahners Business Information. Abstract obtained from Wireless Week, 1+.

[4]Dunnavant, S.T. (2001). Blackberries in support of

technology, Proceedings of the 29th annual ACM SIGUCCS conference on User services, Portland, OR, 36-39. Retrieved from:

[5]Fox, A., Gribble, S.D., Chawathe, Y., & Brewer, E.A.

(1998). Adapting to network and client variation using infrastructural proxies: lessons and perspectives, IEEE Personal Communications, 5(4), 10-19.

[6]Merdich, E.A., Roizen, J., Rubin V., & Buckley, S. (1997).

The serious business of growing up. (Berkeley: University of California Press).

[7]Ghinea, G. & Thomas, J.P. (1998). QoS Impact on User

Perception and Understanding of Multimedia Video Clips, ACM Multimedia 98 - Electronic Proceedings.

[8]Gimson, H. (2002, December 13). Delivery Context

Overview for Device Independence, W3C Working draft. [9]Hsi, S. (2002). The Electronic Guidebook: A study of user

experiences using mobile web content in a museum setting, International Workshop on Wireless and Mobile Technologies in Education, (WMTE 2002). August 29-30, 2002. V?xj? University, Sweden.

[10]Huizinga, D.M. and Heflinger, K.A. (1995). Two-level client

caching and disconnected operation of notebook computers in distributed systems, Proceedings of the 1995 ACM symposium on Applied computing, Nashville, TN, 390-395.

[11]Kindberg, T., Barton, J., Morgan, J., Becker, G., Caswell, D.,

Debaty, P., Gopal, G., Frid, M., Krishnan, V., Morris, H., Schettino, J., Serra, B., & Spasojevic, M. (2002). People,


places, things: web presence for the real world, Mobile

Networks and Applications, 7(5), 365-376.

[12]Mooney, Elizabeth V. (2000). International markets will

present opportunities for fixed wireless [Electronic version].

RCR Radio Communications Report, 19(27), 37.

[13]Penton Publishing, Inc. (2000). Large wireless data market is

emerging [Electronic version]. Microwaves & RF, 39(5), 23.

[14]Stronck, D.R. (1983, April). The comparative effects of

different museum tours on children’s attitudes and learning.

Journal of Research in Science Teaching, 20(4), 283-290.

[15]Wu, K.L., Yu, P.S., & Wolf, J.L. (2001). Segment-based

proxy caching of multimedia streams, Proceedings of the

tenth international conference on World Wide Web, Hong

Kong, 36-44.