Customizing Multimedia with Multi-Trees

Introduction

The majority of multimedia applications rely on hypermedia technologies, such as HTML, XML, or PHP (cf. Lang, 2005, for a review on design issues of hypermedia systems). These technologies enable the presentation of any content such as entries in a digital encyclopedia or products on a company’s homepage. In contrast to database queries, the hypermedia has to be navigated interactively. The navigation process frequently fails, and the user gets lost in hyperspace. This widespread phenomenon (Shneiderman & Plaisant, 2005) is caused mainly by an inadequate navigational design of the hypermedia. Making up an adequate navigational design becomes even more challenging if groups of users differ with respect to their knowledge of a topic’s structure and if they have overlapping interests.

The navigational design comprises two components: the structure of the hypermedia and the layout of user interfaces. The latter aspect is the focus of usability studies (e.g., Falk & Sockel, 2005); whereas, the former is less frequently discussed in the literature and is given scent mention in lectures at universities or business schools. This article is mainly devoted to the former aspect, and:

• •

  outlines the graph theoretic foundations for structuring hypermedia,

• •

  introduces multi-trees for customizing hypermedia with respect to different user groups, and

• •

  provides an overview of metrics to assess the navigational efforts of the user.

The approach presented herein differs from well-established human-computer interaction studies (e.g., Arroyo, Selker, & Wei, 2006), because it aims at quantifying the users’ navigational efforts with respect to the structure of hypermedia systems rather than the interface design. This article presents a modeling approach, and all results are derived by a deductive analysis.

The remainder of this conceptual article is structured as follows: subsequently, the opportunities of structuring hypermedia are outlined. Then components of users’ navigation efforts are discussed, and metrics for the assessment of navigational burdens are presented. Afterward, advantages of multi-trees are highlighted using a numerical example. Starting from a discussion of the limitations of this approach, avenues of future research are pinpointed. The final section provides the conclusions of this study.

Structuring Hypermedia

Hypermedia are networks comprising media objects (documents, pictures, films, etc.), pseudo-objects (pages for guiding the user), and links to interconnect media objects and pseudo-objects. In terms of graph theory, both media objects and pseudo-objects are nodes (or vertices) of a graph, which provide some content or navigational information for the user. The links are the edges of the graph, which enable navigation of hypermedia. Since the nodes are arbitrarily types of media (films, sounds documents, etc.), this simple organization scheme holds for many multimedia instances of our everyday life, of which the World Wide Web is clearly the most prominent.

For the design of adaptive hypermedia two types of nodes are distinguished (Brusilovsky, 2001; Muntean, 2005): navigational nodes and contents comprising nodes. If different types of media are knitted within one environment, the navigation within the media-objects has to be considered as well as the navigation between them. The subsequent considerations are restricted to the latter problem of reaching the individual media objects of interest, but do not address the complexity arising from the combination of different qualities of media objects.