Hypermedia Application Design: Guidelines and Case-studies in the Law Domain

• Autore: Franca Garzoto, Luca Mainetti, Paolo Paolini
• Pagine: 13-30

Page 13

@1. Introduction

Hypermedia applications include static media such as text and graphics, as well as active media such as video, animation and audio, coupled with sophisticated navigation capabilities similar to those found in hypertext applications. Designing hypermedia applications that are significantly larger than a few dozen nodes and links is a complex task, if compared to other, more traditional classes of information systems. Complexity is induced by a number of factors: the non-linear organisation of the material and the need, for the designer, of controlling the potential "explosion" of the number of links; the co-existence of multiple media of substantially different nature, that induce a variety of styles of user interaction; the lack of standardisation, both in the architecture of hypermedia systems, and in the definition of information structures and interaction paradigms that should be provided in hypermedia applications.

In this paper, we will analyse the process of hypermedia application design, and we will propose some guidelines to address the design task in an organised way. According to our approach, the design of a hypermedia application can be organised in four different (but interdependent) tasks: content analysis, structure design, dynamic design, and lay-out design [7], Each of these tasks focuses on different aspects of a hypermedia application and requires different methods, description terminology, and expertise, in this paper, we will also propose a set of primitives that can be useful for describing the output of the design process and are based on a well-known design model called HDM [1, 3, 4, 5].

Our approach will be exemplified by discussing the design of a hypermedia in the law domain. This case study application is named CLICK and is commercially available on a cd-rom. It is produced and distributed by an Italian publisher specialised in economic and legal publications [2].

Page 14

@2. Hypermedia Content Analysis

Content analysis is the task of selecting and producing the actual information that must be included in the hypermedia application. The content can be represented by static, passive media such as formatted data, text, imagef or by dynamic, active media such as video clips, sound tracks, animation, etc.

This design phase requires one or more specialists in the application domain, supported by an application analysts, in order to define the user requirements and the intended goals of an application. In this design phase, it must also be decided if the material must be created ad hoc for the specific application, or If pre-existing material can be reused.

Content analysis results in the specification of the knowledge requirements of the application and in the identification of the source material and of the methods to produce it.

The organisation of the content and the specification of dynamics and lay-out features of a hypermedia application are concerns of the design phases, discussed in the following sections.

Example

Our case study "click" [2] concerns Italian fiscal regulation, and is mainly intended for professionals such as fiscal specialists or lawyers. The click CD-ROM stores the laws regarding Italian fiscal regulation from 1924 to 1994¹, and their relationships to the related ""practice" (I.e., law enforcement) and case law. This material Is integrated with comments and argumentations that discuss legal Interpretation In order to orient the users In the application of the various laws, click partially reuses the material already pulished in text books and manuals by the publisher, updated with the text of- the most recent laws and their related case law and practice.

@3. Hypermedia Structure Design

Structure design refers to the way information is organised within a hypermedia application. Structure design requires the ability of abstracting the general organisation patterns from the analysis of the application content and user requirements. This design phase also requires a structural design model, I.e., a set of primitives to describe precisely and concisely the structu- Page 15 ral aspects of an application, Structure design of a hypermedia application results in the specification of a set of structure types, at two levels, in-the-small and in-the-large. In the following sections, the description of the structural design will be discussed using the terminology of the HDM Model.

Further information about this hypermedia design model can be found in references [1, 3, 4, 5].

@@3.1. Structure Design in-tbe-Small

Hypermedia structures in-the-small are called nodes and slots in HDM terminology. A node represents a semantic unit that can be regarded as atomic, i.e., it cannot be further decomposed. A node aggregates atomic or composite pieces of information that are called slots and are made of simple or composite values of different nature (text, formatted data, images, sound,..) [9].

The visualisation of slots within a node, their use, and their interaction with the user, concerns dynamic and lay-out design and will be discussed in sections 4 and 5.

Structure design in-the-smai results in a set of in-the-small structure types, i.e., slot-types and node-types, Slot-types can be simple denoting multimedia data-types such as "text", "video-clip", "still", or can be com-posite denoting aggregations of simple or in turn composite slots.

A node type defines a class of nodes that aggregate slots of the same type, and represents semantic units of the same "class". Examples of node types in our case study applications will be discussed in section 3.3,

@@3.2. Structure Design in-the-Large

In-the-large structures are "large" granules of information that aggregate several nodes and represent complex objects or relationships of the application domain, HDM provides two primitives for in-the-Jarge structures, called collections and webs.

A collection represents a physical or conceptual entity of the application domain, or a set of information structures - node or collections - that are aggregated for a specific purpose (e.g., because they are concerned with the same general topic) [6]. A web is used to represent a relationship between a node and another node, or between a node and a collection.

A collection is made up of a node called collection node, a set of members, and a topology, that defines how members are organised within a collection. Typical topologies for collections are sequences and trees. Different topologies induce different navigation patterns, as we will discuss in section 4.

Page 16

Collection members can be nodes, or other collections in turn. In the latter case, collections will be called nested. The collection node has the purpose of introducing the collection, and of allowing access to its members. Very often, its content is (partially) derived from the collection members, in that it reuses some slots occurring in the nodes of the members.

Structure design in-the-large of a hypermedia application results in the specification of a set of collection types and a set of web types.

Collection types define classes of collections representing similar objects, or sets of objects that are aggregated with similar puiposes. Collections of the same type all share the same topology, and have collection nodes and members of the same types. A web type defines- a class of relationships that hold among nodes or collections of given types.

@@3.3. Examples of Structure Design

The discussion of structure design in click will proceed top-down, from the analysis of slot and node types (structures in-the-small) to the description of collections and webs (structures in-the-large).

Slot Types in Click

In click, simple slot types denote strings of text, positive numbers or dates. Examples of aggregated slot types are:

* "title" made up of the composite slots "identification", "subject", and "general topic"; "identification" is in turn made of the slots "regulation type" (e.g., Law, Decree, EC Regulation etc.), "number" and "pulication date";

* "paragraph", that represents a portion of the plain text of a legal document or an article of a law;

* "annotated paragraph" made of a paragraph slot and a set of additional text slots that either comment on its interpretation or list some related laws;

* "paragraph with versions", made of an annotated paragraph and a text...