Novel SVG Visualizations for Exploring Distributed Digital Libraries

Keywords: Digital Library, Novel Visualization, Algorithm, Web Services

Qianyi Gu
Ph.D Candidate
Department of Computer Science, University of Colorado at Boulder
Boulder
Colorado
USA
qianyi.gu@colorado.edu

Biography

Qianyi Gu is a Ph.D. student at the University of Colorado at Boulder, Department of Computer Science. He works with the Boulder Learning Technologies Group under the leadership of Dr. Tamara Sumner. His research topics are artificial intelligence, human-computer interaction and semantic web and search. He received his M.S. degree from the Computer Science Department, State University of New York at Stony Brook. He is currently involved in several digital library research projects: the Digital Library for Earth System Education (DLESE www.dlese.org), and the National Science Digital Library (NSDL).

Faisal Ahmad
Ph.D Candidate
Department of Computer Science, University of Colorado at Boulder
Boulder
Colorado
USA
fahmad@colorado.edu

Biography

Faisal Ahmad is currently a Ph.D. student at University of Colorado at Boulder. He works with the Boulder Learning Technologies Group under the leadership of Dr. Tamara Sumner. He has served as student chair on the 2005 Joint Conference on Digital Libraries. Current projects involve enhancing the usage of digital libraries by linking resource discovery and educational standards and modeling and developing knowledge organization services. His other interests include ubiquitous computing and educational technologies.

Dr. Francis Molina
Technology Director
Project 2061, American Association for the Advancement of Science
Washington D.C.
USA
fmolina@aaas.org

Biography

Francis Molina holds a PhD in Botany from the University of British Columbia, Canada. He worked for 8 years as a research scientist at the American Type Culture Collection where he performed molecular characterization and computer-assisted image analysis and identification of microbial strains. His interest in computers led him to pursue a certificate in Interactive Multimedia and Web Development at the George Washington University. He is the technology director of Project 2061, a long-term education reform initiative of the AAAS (American Association for the Advancement of Science) where he introduced SVG as a key technology for developing interactive strand maps which depict students' progression of understanding of key science ideas.

Dr. Tamara Sumner
Associate Professor
Department of Computer Science, University of Colorado at Boulder
Boulder
Colorado
USA
sumner@colorado.edu

Biography

Tamara Sumner is Associate Professor in the Department of Computer Science, University of Colorado at Boulder. Her research includes human-computer Interaction, user interface design, usability and education technology. She is Principal Investigator (PI) or co-PI on several NSDL projects and is the former chair of the NSDL Educational Impact and Evaluation Committee.


Abstract


The Strand Map Service (SMS) provides educators and learners with conceptual browsing interfaces (CBI) that help them to locate and use learning resources in educational digital libraries. Different use scenarios and tasks present different concerns and constraints to users of educational resources in digital libraries, necessitating information to be presented in a variety of perspectives. To fulfill these requirements, the service described in this paper introduces a methodology for generating novel visualizations based on the same information space.

The Conceptual browsing interfaces that we are developing have dual use: 1) to help users understand the learning objects and their relationships by providing visualizations; and 2) to facilitate the formulation of search requests for educational resources related to learning goals and the retrieval of these resources from distributed digital libraries. For forming search requests and retrieving resources, the visual interfaces serve as a single, unified gateway for users to explore distributed resources. By interacting exclusively with our visual interfaces, the user can search directly or indirectly for distributed web resources residing in different digital libraries such as the National Science Digital Library (NSDL), the Digital Library for Earth Science Education (DLESE), the Harvard-Smithsonian Digital Video Library (HSDVL) and even in a larger domain such as Google. The retrieval process is automatically handled by services provided through the visual interfaces. Instead of forming separate search queries on different digital libraries, the user only needs to interact with the learning goals found on our visual interfaces, an approach that is very similar to "federated searching". To achieve this goal, we bind our SVG visual interfaces to web services provided by various digital libraries.


Table of Contents


1. Introduction
     1.1 Benchmarks, Strand Maps and the Strand Map Services
     1.2 Concept Browsing Interface (CBI) and Visualization System
2. Novel Visualization
     2.1 Significance
     2.2 Novel Visualization Model
3. Web Service Integration
Bibliography

1. Introduction

1.1 Benchmarks, Strand Maps and the Strand Map Services

Benchmarks are learning goals which describe what learners should know, or be able to do, at key stages in their primary and secondary education across the natural sciences, mathematics, technology, and social science disciplines. In the United States, the benchmarks provide a common framework for understanding the relationships between national, state and local education standards. [5] Strand maps are constructed based on the benchmarks articulated in Benchmarks for Science Literacy[3]. They provide a visual representation that emphasizes the coherence intended in the benchmarks and encourage both teachers and learners to make connections between ideas.

The Atlas of Science Literacy[4] , published by American Association for the Advancement of Science (AAAS) and the National Science Teachers Association, features strand maps on topics that are important to science literacy. Each map consists of node-link representations illustrating a set of relationships between benchmarks organized around a topic. High-level descriptions of the benchmarks are provided in the nodes, while the links depict the interrelationships between benchmarks. Each map contains vertical strands reflecting key ideas in that topic. Each strand is cross-referenced by grade levels (K-2, 3-5, 6-8, 9-12) to illustrate how student understanding develops over time.

The Strand Map Service builds on and extends the significant knowledge base embodied in Benchmarks and the Atlas. The Service supports the needs of K-12 (primary and secondary school) educators and learners, and digital library developers through the provision of graphical conceptual browsing interfaces. It also supports the programmatic web service interface which enables digital library developers to easily construct conceptual browsing interfaces appropriate to the needs of their specific library audiences using dynamically generated visual components provided by the Service.

1.2 Concept Browsing Interface (CBI) and Visualization System

The Strand Map Service provides educators and learners with concept browsing interfaces (CBI) that help them to locate and use learning resources in educational digital libraries [5]. The CBIs are constructed based on nationally recognized education standards. Their graphical representations help learners and educators to understand the learning objects and their internal relationships. To achieve these learning goals, the user needs to retrieve educational resources aligned to those objects and use the resources effectively. We have introduced a system that dynamically generates visualizations of these interfaces using SVG [6]. Because of its versatility, SVG has proved to be the ideal format for the SMS. In this paper, we will show our subsequent work on how the CBI:

1. presents different views of the concept space to meet users' requests and different service scenarios by generating novel visualizations from the same data repository; and

2. integrates with different web services to help the users retrieve educational resources from distributed digital libraries.

2. Novel Visualization

2.1 Significance

By presenting benchmarks and their inter-relationships with semantic constraint, the strand map visualization model helps the user understand the dependencies of science concepts on a specific topic. The visualization model:

Novel visualizations are necessary because of constraints in a standard or pure strand map visualization model. Even if it is dynamically generated, a standard strand map organizes and then presents benchmarks from the concept space only if they are part of the same map. A user might have other information needs that are not met by a standard strand map presentation, as illustrated in the following scenario:

A high school teacher is teaching a course in earth science. In a particular lesson she would like to teach the concept (or benchmark) that "Some changes in the earth surface are abrupt while other changes happen very slowly". She goes to the strand maps to find out what other concepts are closely related, either as a pre- or post-requisite to this targeted concept. However, because the same concept might occur in different maps, this means that she has to find all of its instances to see what other benchmarks might contribute to or depend on it. Even if a search mechanism were built to search across the collection of strand maps, they still have to be presented and then examined individually for the dependencies.

Because manual inspection of benchmark dependencies can be a very time-consuming process, a novel, nearest-neighbor visualization is needed so that related concepts from other topics (strand maps) can be displayed with the targeted concept to facilitate its instruction.

2.2 Novel Visualization Model

In this section we describe the different modules that are involved in generating the novel visualization.

User Inquiry Module: Includes the user interface and query generator.

Data Collector Module: Retrieves the data from the back-end database and represents it according to the object model of the novel visualization module. Details of this module are described in our previous SVG paper [6]. In brief, this module:

Visualization Generation Module: This module generates the novel visualization for the user. Some components of this module have been previously described [6]. These components serve to:

The key component for generating novel visualizations in the Visualization Generation Module is the Visual Component Hub (VCH). As discussed in our previous research, the visualization algorithm can dynamically generate a graphical representation of concepts and their relationships [6]. Novel visualization is based on the successful retrieval of the requested visual components and their organization or assembly into a coherent interface using our visual algorithm. The VCH includes several units:

The workflow and interaction between these units is shown in Figure 1. Figure 2 shows the resulting novel visualization in our user scenario.

workflow.svg

Figure 1: Novel visualization system overview

related.svg

Figure 2: Novel visualization of nearest neighbor benchmarks

3. Web Service Integration

Recent studies [1], [2] have shown that educational value of digital libraries can be enhanced by using strand maps for navigation and resource exploration. Students using the strand map interface engage more with science content compared to the those using a keyword-based searching interface. Another finding is that the educational benefit of strand maps diminishes when students move from strand maps to the resource list resulting from keyword searching or clicking at a concept in the strand map. This can be attributed to the disappearance of the context that was provided by the strand maps. To overcome this issue we have embedded the digital library resources in the strand maps so that the user (e.g., a student) presented with a strand map can see the digital library resources within the context of the strand map. Figure 3 shows a portion of a strand map along with the retrieved resources.

example.svg

Figure 3: Digital Library for Earth System Education resources embedded in a strand map

In order to enable flexible integration of digital library resources into the strand maps we have adopted a web-service composition approach and use SVG post-processing. The web-service composition architecture is shown in Figure 4. The composition process proceeds as follows:

1. The user opens a digital library web site.

2. The user clicks on a link to opens a strand map (e.g., Weather and Climate).

3. The digital library uses the SMS client library to send a request to retrieve the target map from the SMS server.

4. The retrieval request also defines which query string should be attached to each concept in the strand map.

5. SMS server constructs the map and then attaches the requested query string, as a hyperlink, to each concept within the SVG.

6. Once the SMS client library receives the map it parses the SVG and uses the embedded hyperlinks to retrieve search results e.g. Google, NSDL, DLESE etc.

7. The SMS client library then parses the search result for each concept and adds those results in the strand map SVG.

8. Finally, the search result embedded strand map is returned to the user/client.

webServices.svg

Figure 4: Web-services composition architecture

The composition process illustrates some of the benefits of using the SVG format and a web-service composition architecture. First, the SVG document can be post processed to add graphics and modify behavior. In our scenario we are dynamically extracting information from the SVG (Step 6), adding content, and changing behavior (Step 7). Second, the SMS client library is interacting with two web servers (the SMS server and the digital library server) to first retrieve the strand map and then to retrieve the search results for each concept in the strand map. The flexible nature of web-services and the SMS client library allows different digital library web services to be used with strand maps and for search results from more than one digital library to be embedded into the strand maps. In order for the web service composition architecture to function efficiently the strand map must be capable of being modified once it has been retrieved from the SMS server. Because of its textual format (vs. binary formats such as Flash), SVG allows for customized post-processing. Thus, SVG plays a critical role in achieving a clean separation between client and server and balancing the distribution of intelligence between them. In conclusion, SVG possesses important characteristics that enable layered software design and a web service composition architecture.

Bibliography

[1]
K. R. Butcher, S. Bhushan, and T. Sumner, "Multimedia Displays for Conceptual Search Processes: Information Seeking With Strand Maps," Submitted to ACM Multimedia Journal, 2005.
[2]
S. Bhushan, "Language for describing digital library components", Unpublished masters thesis, University of Colorado at Boulder, 2004.
[3]
AAAS, Benchmarks for Science Literacy. Project 2061, American Association for the Advancement of Science, Oxford University Press, New York, 1993.
[4]
AAAS, Atlas of Science Literacy. Project 2061, American Association for the Advancement of Science, and the National Science Teachers Association, Washington, DC, 2001.
[5]
Tamara Sumner, Faisal Ahmad, Sonal Bhushan, Qianyi Gu, Francis Molina, Stedman Willard, Michael Wright, Lynne Davis, and Greg Janee, "A Web Service Interface for Creating Concept Browsing Interfaces" D-Lib Magazine, November 2004, Volume 10 Number 11
[6]
Qianyi Gu, Faisal Ahmad, Francis Molina, Tamara Sumner, "Dynamically Generating Conceptual Browsing Interfaces for Digital Libraries Using SVG", 3rd Annual Conference on Scalable Vector Graphics, Tokyo, Japan , Sept 7-10, 2004

XHTML rendition made possible by SchemaSoft's Document Interpreter™ technology.