|Table of Contents|
|Table of Contents|
In the typology of the information, it is straightforward to incorporate an appropriate domain-oriented characterisation in the model. For example, INSPEC terms can be used in physics; for chemistry they can be replaced by the terms defined for Chemical Abstracts. An appropriate bibliographic component can also be chosen for any domain. The set of bibliographic data we have used to characterise the information can be extended and refined in a straightforward way. We have not addressed the question as to what would be the most appropriate way of labelling modules by their domain-oriented and bibliographic characteristics; further research may be required.
The range component can be useful in any domain in which research is organised in projects and programmes with a shared background. Our characterisation by range can be refined, but it is also possible to replace our range component with a binary characterisation that distinguishes only between regular articles and `background modules'.
The component of the conceptual function is the main component of the typology of the information. The set of six main modules that we distinguish by the conceptual function can be used in areas of research with a problem-solving structure in which, using a particular method, results are generated and then interpreted. In other words, the basic modular structure with respect to the conceptual function may be applied to experimental and empirical research. In the physical sciences, experimental results are explained in terms of natural phenomena; in other fields the interpretation may be in terms of sociological, economical or other phenomena. Within these main modules, more and other constituent modules can be defined, to tailor the modular structure to a particular domain.
The typology for the links is less domain-specific than the typology for the modules. It has two main components: one for expressing organisational relations and the other for scientific discourse relations. Each of these components is subdivided into a hierarchy of smaller components. The organisational relations we have defined can easily be identified and expressed in any modular publication created using a modular model that has a range-component.
The component of the scientific discourse relations has two main subcomponents: content relations and relations based on the communication function. Concerning the content relations, characterised links expressing dependency relations in the problem-solving process and those expressing elaboration relations can play an important role in publications of many genres in many domains. It is useful only to include our rather elaborate component for the communicative function in a modular model for a domain and genre in which different communicative functions (such as argumentation, explanation, clarification) are aimed for explicitly. The scientific discourse relations are also the most difficult to make explicit in links and the link typology remains open for further research. In particular, current research in discourse analysis and argumentation theory warrants a more elaborate analysis of the scientific discourse relations.
Our modular model is intended for the creation and evaluation of modular articles. It can be extended with a component dealing with comments on modular articles. Such an extension could be based on Trigg's typology for `comment links' [Trigg, 1983].
Based on the typologies for the identification of information units and relations, specific instructions must be formulated for the creation and evaluation of the different types of modules and links in which respectively the information and the relations are represented. A module that requires particular attention is the module Abstract , which fulfils an important role in the orientation of the reader. In this thesis, we have not addressed the question of how the abstract of a modular article should be structured and linked to the body of the article in order to fulfil this role. The abstract is the subject of a separate thesis within our project (see [Van der Tol, 1999]).