Humanities are becoming a hyping field of intensive research for computer researchers. It seems that conceptual models may be the basis for development of appropriate solutions of digitalisation problems in social sciences. At the same time, humanities and social sciences can fertilise conceptual modelling. The notion of conceptual models becomes enriched. The approaches to modelling in social sciences thus result in a deeper understanding of modelling. The main aim of this paper is to learn from social sciences for conceptual modelling and to fertilise the field of conceptual modelling.
1.1 1.2
Digital humanities is becoming a hyping buzzword nowadays due to digitalisation and due to over-applying computer technology. We have been engaged in a number of projects, e.g. 1.3
These observations and lessons are useful for conceptual modelling in our area. They are mostly not explicitly observed in computer science. They are however implicitly used. Think, for instance, on conceptual database models. We often use a conceptual schema that describes the structure of the entire database system and use additionally a number of conceptual views that describe the viewpoints of users. Therefore, we explain now how conceptual modelling can learn from successful approaches in social sciences. The learning process will enhance the added value of conceptual modelling. 3 The mould is a hollow form or matrix or simply frame for giving things (such as models) a particular shape. In production, moulds are used as a shaped cavity for forming fluid or plastic things.
2
According to [
A notion of conceptual model might be a slim, light, or concise one depending on the level
of detail we need in model utilisation. We will use in the sequel one notion, i.e. the concise
notion and refer for slim and light versions to [
Conceptual Model w (Model É Concept(ion)s) ./ Enabler [
A conceptual model is a model that is enhanced by concept(ion)s from a
concept(ion) space, is formulated in a language that allows well-structured
formulations, is based on mental/perception/situation models with their embedded
concept(ion)s, and is oriented on a mould and on deep models that are
commonly accepted.
The mould and the deep models form the matrix of a model [
Models do not have to be conceptual models. Conceptual models do not have to be based on an ontology. The main purpose of conception as a system of concept or of a concept(ion) space is the integration of interpretation pattern that ease the communication, understanding, delivery of a model in dependence on the model functions. The concept(ion) space, the mould of model utilisation, and the explicit knowledge of the social determination provide a means for the correct and suficiently precise interpretation of the model elements. 2.3
The consideration of the strategic, tactical, and operational sides of modelling and of
conceptual modelling drives us to consider the four dimensions in Figure 1. These dimensions
cover the application areas in [
dimension.
Prescriptive conceptual models that are used as the blueprint for system realisation also consider this dimension. The social embedding is typical for social sciences. The foundation dimension has additional aspects in social sciences since corroboration, comprehension and systematisation are far more complex. Conceptualisation is based on complex concept and conception spaces. 2.4
In is often claimed that conceptual database or data models are mainly descriptive ones. Description is, however, only one of the functions that a conceptual data model has in a system development scenario. Other typical scenarios are documentation, prescription, communication, negotiation, and explanation. These scenarios are also observed for humanities.
In system construction we transform the conceptual data model to corresponding realisation models. This transformation also changes the semantics from rich semantics of conceptual models to lexical semantics which is based on the lexical interpretation of the words used in realisation models according to the meaning in the given application area. It is thus forgetful. The reestablishment of the conceptualisation must thus be handled by a reference to the conceptual model what also means to use a tight bundling of all models in the case of system maintenance (e.g. evolution and migration) and integration. The social dimension and the foundation dimension get also lost during transformation. 2.5
Based on the observations, we should consider a conceptual model as a model suite, i.e. a
coherent collection of explicitly associated models. The associations are explicitly stated,
enhanced by explicit maintenance schemata, and supported by tracers for the establishment
of coherence [
Model suites in most sciences and engineering incorporate some conceptual models. This
situation is not diferent for social sciences. For instance, the CRC 1266 [
Middle-range theories [
This approach often results in a gap between qualitative and quantitative research. Models can be used to render the theory ofer. At the same time, models may also render a qualitative theory. The rendering procedures are typically diferent. A model suite can now be constructed by models for theoretical concepts from one side and by models for quantitative concepts from the other side. In this case, we use models for the quantitative theory ofers and for the qualitative theories. This approach is depicted in Figure 2. quantitative f¹¹ss;;mqºº / suspopuorOrcteisng o f¹¹ss;;mtºº (supporting sources sou rOces quantitative research g¹s;qº ¹c;qº quantitative conc e_pts
hq¹c;qº theory ofer f¹¹cc;;mqºº ofer driven view
models as mediators f¹¹ss;;tqºº model / conc e_pts o
G¹¹sc;;mmºº h¹mc;mº / model o
qualitative theory-oriented
research f¹¹cc;;mtºº theory driven view
G¹¹sc;;ttºº theoretical conc e_pts
ht¹c;tº
theory
This approach has already used for the investigation in the CRC 1266 [
3
Conceptual modelling is a widely used practice in many science and engineering disciplines. The current practice used for database conceptualisation can be enhanced by a number of insights that we observed in social science research.
The concept(ion) space is often far more complex structured than finally represented and used for a singleton conceptual model. We should consider conceptual models that orient on diferent aspects and diferent levels.
The context dimension should not be neglected for conceptual models.
The social dimension and the foundation dimension are equally important as the representation dimension. form a model suite.
Models and especially conceptual models consist of a number of models and thus We got now additionally a number of special necessities for conceptual modelling without which conceptual models are of low quality, not justified, and also not adequate. Deep models: Models consist of normal sub-models and deep sub-models. The first ones are given in an extrinsic and explicit form. The later ones are often concealed. Model mould: The second element of the matrix of modelling is the mould. We know a number of canonic approaches that guide the modelling process, the modelling outcome, and the capacity of the finally developed model.
Concept-biased modelling: Conceptual models are typically deeply biased by the concepts in a given domain. Concepts such as “village”, “settlement” and “center” are essentially representing the same understanding but are used in very diferent contexts. The same applies to database models, e.g. the concepts of “Person” or “Address” depend on geographic, law etc. assumptions.
functions that a model has in such scenarios. The model is an instrument in these scenarios. Whether it is a proper and fit-to-use instrument depends on the function the model has (and thus on the purpose and the goal).
We thank the reviewers for their remarks, suggestions, and critics. The paper is based on our previous papers
on models, e.g. on the generalisation of approaches used in design science research [
This research was performed in the framework of the CRC 12664 ’Scales of Transformation Human-Environmental Interaction in Prehistoric and Archaic Societies’ which is funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation - Projektnummer 2901391021 - SFB 1266)5. We thank both institutions for enabling this work.