Taxonomies and thesauri

Taxonomy is another concept which has experienced a change in meaning with the onset of the computer age. In this case, the term has experienced an increase in its field of action. The use of taxonomy in its classic form as applied - almost exclusively - to biology and logic has evolved to include its application to electronic technology. Thus the term taxonomy has begun to be used in the context of Information Science, specifically with regard to classification systems (see Figure 2.1).

Figure 2.1 Concept of taxonomy

The application of the term taxonomy to systems used for the classification of subjects is a topic which until now has been discussed by very few specialists in information and documentation. One example of its treatment can be found in my book Documentación y metodología de la investigación científica (Documentation and Methodology of Scientific Investigation), written in 1985, where I refer to the relationship between taxonomy and classification systems. However, the term has been increasingly used over the last few years, above all in conjunction with the concepts of information architecture and mining of information, and always in the context of knowledge organisation in companies and institutions. Above all, the term has been used in the context of semantic webs.

Recently, professionals in Information Science have in fact taken their concept of taxonomy from computer scientists who have found in taxonomy a solution to their classification problems, especially when it comes to the ordering of conceptual units. They extract conceptual units from documents in order to facilitate the construction of computer programs for the indexing and retrieval of information.

From ordering to taxonomy

Taxonomy is, therefore, a very relevant issue in modern times, and it is necessary to examine the concept in its main traditional context in order to understand its subsequent application within the general context of Information Science.

Following a logical process, working from the lowest to highest complexity, means dealing first with ordering, then classification and classification systems, eventually reaching taxonomy. Classification systems include general or universal systems and specific systems based on isolated conceptual units, terms or their components, structured according to certain methods.

Therefore ordering requires a physical process which consists of assigning each entity - whether real or abstract - to a location. If we are dealing with a real entity - such as a book - its ordering means finding it a place on a bookshelf. If we are dealing with an abstract entity - such as a thought, for example - then it will be organised in order to occupy a place in our mind or in written form if we want it to endure. Classification represents a higher level of complexity because it can be defined as an ordering process carried out through the application of a particular method. This method implies a process of comparison of the elements to be classified in order to determine similarities and differences between them, which will in turn determine the position of certain elements in relation to others.

Note: It is convenient to adopt this definition of classification in this case, which is similar to the definition proposed in Chapter 1, but a little more appropriate to our current purposes.

The method to be used, the classification system, is decided on beforehand as a pre-coordinated system that can be used for any kind of entity, whether real or abstract.

The origins of taxonomy

Taxonomy, etymologically speaking, comes from the Greek taxis = ordering and nomos = law, norm, rule. In the available literature on the subject we can also find the terms 'taxinomy' and 'taxenomy', which without a doubt have their roots in the different endings that the original Greek expression would have gained from the different cases of its declension. Alvarado Ballester (1967) defines taxonomy as 'the theoretical study of a system, including its basis, principles, procedures, and rules'. It is also defined as classification theory (see Figure 2.2).

Figure 2.2 Taxonomy in information and documentation science

Thus, once again, the topic moves to a higher level of abstraction. It is no longer sufficient to establish a method of classifying entities (in biology these would be living things or their structures): it is also necessary to carry out a theoretical study which takes any other relevant aspects into account. The term taxonomy was coined by A. De Candolle in 1813 in order to design the rules or laws to be used in systematics. Since then it has been used almost exclusively as a component of biology. Nevertheless, within the field of logic it has a wider meaning.

Using taxonomy it is possible to develop methods established after the fact, once the classificatory elements are known. The determinants of order are not decided beforehand. It is important to take the individuality and specific characteristics of the entities to be ordered into account. A structural unit is decided on, which will also represent the classification unit which Crowson (1970), among others, defines as a taxon.

Using taxonomy it is possible to establish categories within a classification depending on how relationships of similarities (the interaction principle), or relationships of interdependence (the duality principle), will be determined. In the first case, a classification would be achieved in a horizontal direction representing the correlation between the taxa. In the second case, a hierarchy will be used which establishes a scale from greatest to smallest, from superior to inferior entity, which will give a sense of collectivity and generality, in contrast to the previous case where a sense of individuality is achieved.

Hierarchical and horizontal order

In a taxonomy with a hierarchical order, the taxon to be used should be decided on first of all, along with the desired level of subdivision. In the field of chemistry, for example, it would be necessary to determine whether the taxon will be an element, a molecule, an atom, etc., and whether subdivisions above or below are going to be created, in order to subsequently establish the rules and laws of classification. In Information Science the taxon could be:

 a taxon in a conceptual unit

which would be the most logical choice, although it could also be:

 taxon = data or document.

The term conceptual unit (concept unit, expressed by the terms, or parts of terms, of any given language) should not be confused with the idea of a unit of information. A unit of information can be simple or complex and consist of one or more conceptual units. On the other hand, there is no information if it is not perceptible. That is to say, a conceptual unit can exist on its own, but a unit of information cannot: it must be recorded in a receiving subject and, consequently, transmitted by an emitting agent (see Figure 2.2).

In the context of this study, taxonomical methods with a horizontal structure dealing with the correlation between taxa would refer to classifications, facets and keywords. In contrast, great classifications such as that of Dewey, the Universal Decimal Classification and the classification used by UNESCO have a hierarchical, vertical, taxonomic structure.

Correlation with classifications

It is now worth specifying some concepts and details in order to better establish the correlation between taxonomy and classifications.

Durand de Gros (1899) is one of the scientists who have established the basis for taxonomy including:

 order of generality - where the taxa are natural elements which are ordered from greatest to smallest;

 order of composition - where only concrete objects are used to form the taxa.

A relationship between the whole and the part - and vice versa - is established:

 order of genealogy;

 order of evolution - where the taxa are treated according to their origin.

These form the basis for relationships of:

 hierarchy or subordination;

 coordination; and

 polarity.

Taxonomy allows us to consider:

 a system - creating an artificial order (belonging to the group ordered); or

 a method - taking a natural order into account (for ordering real things or objects).

On the other hand, taxonomic groups consist of entities characterised by a certain reality, due to the fact that the taxa used as a starting point have real characteristics. The relationship between these real characteristics is arbitrary and depends on the criteria...