Educational metadata in use

Abstract

The traditional search and navigation techniques of the Web are wholly user-directed. The burden of pruning the search space is placed upon the user, and is a difficult, time-consuming, and often a (mostly) manual task. This can be particularly problematic when the user is under time, or learning and comprehension constraints, as in educational situations. This short paper describes how we have utilized metadata about educational Web resources to improve the search and navigation processes. In particular, we describe our education-related metadata dictionary, the Courseware Description Language (CDL), and its use within two applications: (1) COOL Links, a user-centered hypermedia link type; and (2) SeekEZ, a user-centered searching facility. We also provide a brief survey of the metadata efforts that have been useful to education and instruction.

Keywords

Metadata; Education; Search; Hyperlinks; COOL links

1. Introduction

• The Dublin Core [5], born in the field of digital cataloguing, defines a small, core set of metadata attributes for describing the "document like" features of generic Web resources. Many of the current efforts in metadata specification and standards are based upon the firm foundation of the Dublin Core, providing extensions where the core is not specific enough.
• The Gateway to Educational Materials (GEM) project has extended the Dublin Core with metadata to support the description of lesson plans, curriculum units, and other education resources. GEM is a special project of the Educational Resources Information Center Clearinghouse on Information and Technology (ERIC/IT).
• The Warwick Framework [1] provides an architecture for the interchange of distinct metadata packages. The framework itself is distinct from the syntax or semantics of any specific metadata dictionary, and promotes interoperability and extensibility by allowing tools and agents to selectively access and manipulate metadata packages as a whole.
• Most recently, the Learning Objects Metadata Group (LOMG) of the US National Institute of Standards and Technology (NIST) has developed their own set of attributes for describing "learning objects." In fact, they have their own core set of elements, which they refer to as the Cupertino Core. As a contributor to the Instructional Management System (IMS) effort, one goal of the LOMG is to define a metadata specification for submission to the IEEE P1484 working group.

This short paper describes our approach toward building user-centric education applications using metadata. We first offer a brief description of our own metadata dictionary and its relation to those described above, followed by a description of our current application areas.

2. The Courseware Description Language

We first developed our dictionary in late 1995, independently of the metadata work in digital cataloguing (Dublin Core) and online clearinghouses (GEM/ERIC). Yet, as one might suspect, there are many similarities. The elements of the Dublin Core map directly to elements in CDL. The CDL can be used to describe resources that serve as part of a lesson plan (e.g., a Web page which presents activities for learning about solar energy, appropriate for students in 5th grade), similar to GEM. Finally, and in relation to LOMG/IMS, the CDL provides support for building sophisticated knowledge applications (beyond searching and indexing) by identifying those resources that can most appropriately match the needs of an individual student in a given learning context.

The CDL was developed as one part of a formal and implementational approach to automating individualized lesson plan generation. The primary goal of our work has been to address the individual needs of students, including the special needs of students with disabilities. The CDL is firmly grounded in current theories of learning, educational psychology, curriculum and instruction, and teacher decision-making.

3. Educational applications of metadata

User-centric navigation with COOL links: Since early 1996, we have been working on a multi-ended and computed link type, termed a COOL link (or Collection Of Objects Link). A COOL Link is associated with a collection of component resources, via their metadata characterizations. (See [2] in these proceedings for more information.)

User-centric searching with SeekEZ: Our most recent efforts have been devoted to improving the search process by providing personalized resultant sets. Similar to the metadata search engines that are being built today, our notion of search would likewise key on specific, standardized metadata tags. The difference is the amount of allowable personalization. As with all of our work to date, we assume that a user profile exists and is rather comprehensive in nature, particularly relating to the domains in which the individual typically searches and browses. Our search facility has a taxonomy of resources, similar to what might be found with other search engines such as Yahoo!. Yet, our taxonomy is specifically geared toward K-12 education, and each branch is a COOL link. Therefore, as the learner selects a node in the tree, the remaining taxonomy is dynamically constructed by comparing the metadata of the sub-branches an leaves with the learner's profile. COOL links, in effect, are dynamically pruning the search tree as the user directs the query.

4. Conclusion

We believe that the continued success of the Web is contingent upon automated tools that efficiently guide the information gatherer toward relevant and appropriate material. In this paper, we have described our use of educational metadata in building applications for personalized navigation and search in the context of learning. Ultimately, we believe that the Web will be a viable environment for providing individualized instruction applications for parents, teachers, and learners.

Acknowledgments

This research and development was supported (in part) by DARPA contract #N66001-95-C-8626, and US Department of Defense Education Activity (DoDEA) contract #N66001-97-M-C528.

References

URLs