Table of Contents  
Index  
Glossary  


Table of Contents  
Index  
Glossary  
The most obvious aspect by which information in scientific articles is characterised is from a domainoriented point of view, by its contents. Using this characterisation complementarily to the characterisation by the conceptual function given in the previous section, we can distinguish, for instance, between different Treated results modules focusing on different reactions.
The domainoriented characterisation is generally implemented in index terms, such as key words or classification codes. The index terms can be uncontrolled terms invented by the author, or controlled terms that are derived from a classification certified and maintained by some indexing authority. The characterisation space of this aspect of the information is spanned by at least one dimension, with a variable corresponding with the elementary scientific concepts. But for an adequate representation, the conceptual space should be spanned by more than one dimension, allowing for a complex domainoriented characterisation using different key words. The number of dimensions spanning the space is not limited a priori, but dependent on the requirements for an adequate representation.
We do not include in the modular model a detailed characterisation space for the domainoriented characterisation. The development of a suitable characterisation of information by its subject is the key issue in the field of Information Retrieval.^{4.16} With respect to the geometry of a domainoriented characterisation space, in the area of Information Retrieval, `concept maps' are developed with a geometry and a sensible notion of distance [Salton, 1983]. Therefore, we assume that there will be a suitable domainoriented typology available to characterise the scientific content of the information. (As an illustration, some of the terms we used are given in table 4.1.)