The Behavioural Pattern Description Language is designed to enable various types of analysis, based on the transitions in a creativity map. States and their values cannot be described with the description language. The approach has left out states on purpose, as this would raise the complexity of the description language. Otherwise, it could not satisfy one of the main requirements: Simplicity. Another problem arises, if a certain viewpoint should be hidden in order to be able to analysis the correlation of transitions belong- ing particular viewpoints. The BPDL itself provides no operator for actively excluding transitions from the information extraction process.
Fortunately, there is an efficient solution for both cases. The in Chapter 3 discussed partial creativity map (PCM) [130] allows the refinement of creativity maps through the hiding of certain information like viewpoints or by state-based criteria. In Chapter 3, the reader has learned that a state is not only representing the artefact, but storing a set of information. The information is stored in a so-called state vector. Hence, the criteria can be one or more entries of the state vector.
Analyses, based on transitions and states, require two processing steps. The first step includes the creation of the PCM with a suitable hiding condition. The second step is the application of one or more patterns on the PCM.
Following is an example of a creativity map. The creativity map belongs to a project of a software developer. The example will demonstrate the creation of a partial creativity map, which hides the transitions of one viewpoint.
discussing coding coding thinking reading coding discussing reading thinking coding thinking testing testing testing
Figure 4.18: A Creativity Map
Four viewpoints can be identified in Figure 4.18. These are:
Creationwith <−→coding> ,
Knowledgewith <−→thinking, −→reading>
Testwith <−→testing>,
and Needwith <−→discussing>
The creator might want to analyse the creativity map without the viewpoint Creation.
This will result in a new Partial Creativity Map P :
< P, −→P> = < T, −→T> × ( [ i < Ai, −→Ai> \ [ j < Hj, −→Hj>)
where T is the temporal viewpoint, A the set of all viewpoints and H the set of hidden viewpoints (H = < Creation, −→Creation>):
The PCM is depicted below. All four coding transitions have been removed. discussing thinking reading thinking discussing reading thinking testing testing testing
Figure 4.19: A Partial Creativity Map
The creator can now extract the information he requires. The following expression is used to extract a sequence from the PCM.
< pattern > :=
< c transition1 >< c transition2 >;
with < c transition1 >:=“thinking” and < c transition2 >:=“testing”.
The pattern is defined as a thinking transition, followed by a testing transition.
thinking testing
2x
Figure 4.20: Results of the Extraction Process
The result of the extraction process is depicted in Figure 4.20. The sequence has been found twice inside the PCM. It is important to note that one of the sequences would not have been found if the expression had been applied on the normal creativity map. The reason for this is that the coding transition on the bottom of the figure has been hidden.
An analysis of a PCM, based on an entry of the states, would be identical to the example above. The presented approach enables a huge variety of analyses. It includes states into the analysis process as well as the interaction of states, belonging to one or more viewpoints.
4.9
Summary
The first pages of this chapter provide information about the scope of the thesis. It is explained that the presented research is part of the Creative Technologies Research Programme at the Software Technology Research Laboratory (STRL). The creation of the creativity maps is not part of this thesis, but of another. Hence, it is always assumed that the techniques for the creation of those are already provided.
The concept of behavioural patterns is introduced in this chapter, in order to be able to characterise creative behaviour within the creativity map. The number and structure of behavioural patterns is not limited. The approach represents an open technique which can be adapted to the needs of the person who is analysing the creativity map. A behavioural pattern is therefore not only used to describe a creative behaviour, but also supports the search for creative behaviour.
This approach is the solution for research question number 1: How can information about creative behaviour, which is hidden inside the creativity map, be characterised?
Even though, it is possible to analyse the creativity maps by hand. The real strength of the approach emerges with a computer-supported analysis process. A precondition for a computer support is the ability to describe the behavioural pattern in a structured and computer-readable way. This requires a flexible and reliable solution. A description language is introduced in this chapter as an elegant and efficient solution.
The so-called Behavioural Pattern Description Language (BPDL) provides all necessary facilities for a successful application of behavioural patterns. The chapter provides detailed information about the structure of the description language. The description language uses a EBNF style notation. Descriptions of behavioural patterns can be defined quickly and easily. This satisfies the requirement for a simple, flexible and reliable solution. The language is not limiting the number of patterns. New patterns can be described easily. As a result, even people with no knowledge of programming languages are able to define behavioural patterns with the BPDL.
Some of the most important types of behavioural patterns have been explained with ex- amples. The parsing and extraction process works with normal creativity maps as well as with partial creativity map. The latter type makes it possible to refine the input for the parsing process. This allows an analysis, based on states, even this is not directly provided through the behavioural pattern description language.
The architecture of a parser for the Behavioural Pattern Description Language is described in Chapter 7. It represents the second part for answering research question number two: Which computer-aided technique enables the extraction of information from creativity maps?
Knowledge Management and Store
Objectives
Introduction of a knowledge repository for the management and store of the be- havioural patterns.
Description of the content of the knowledge database.
Description of the functionality of a knowledge management facility. Introduction of techniques for coping with synonymy of terms.
“It is beyond a doubt that all our knowledge begins with experience.”
5.1
Introduction
Behavioural patterns are based on the experience and knowledge about creative processes. New knowledge is derived from the already known in a constant process. This knowledge is certainly not only the knowledge of a single person, but of many people from the same or other domains. It is necessary to store the knowledge about creative processes and creativity in a central database. The database must enable the management of all the behavioural patterns.
This chapter is presenting a concept for managing knowledge, consisting of behavioural patterns and their associated information. The facilities of the framework enable the access, adding and editing of the knowledge entries. It is also possible to autonomously search and compare behavioural patterns of different users. Through this, is is possible to identify common behavioural patterns. The content of this chapter represents the solution for research question number three.