Premium
Clustering: An Essential Step from Diverging to Converging
Author(s) -
Tassoul Marc,
Buijs Jan
Publication year - 2007
Publication title -
creativity and innovation management
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.148
H-Index - 60
eISSN - 1467-8691
pISSN - 0963-1690
DOI - 10.1111/j.1467-8691.2007.00413.x
Subject(s) - cluster analysis , gestalt psychology , computer science , process (computing) , object (grammar) , context (archaeology) , fuzzy clustering , artificial intelligence , epistemology , geography , philosophy , archaeology , perception , operating system
Within the context of new product development processes and the Creative Problem Solving (CPS) process, the authors have come to the view that clustering is to be seen as a separate step in the process of diverging and converging. Clustering is generally presented as part of the converging stages, and as such categorized as a selection technique, which in the authors' view does not do justice to this activity. It is about expanding knowledge, about connecting ideas, and connecting ideas to problem statements, functionalities, and values and consequences. It is about building a shared understanding, in other words about ‘making sense’, an essential creative activity in the development of concepts and, although different from a more freewheeling divergent phase, can be as creative and maybe even more so. Four kinds of clusterings are distinguished: object clustering, morphological clustering, functional clustering and gestalt clustering. Object clustering is mainly aimed at categorizing ideas into an overviewable set of groups of ideas. No special connections are being made, other then looking for similarities. Morphological clustering is used to split up a problem into subproblems after which the ideas generated are considered as subsolutions which can then be combined into concepts. Functional clustering is interesting when different approaches can be chosen to answer some question. It permits a more strategic choice to be made. Gestalt clustering is a more synthesis like approach, often with a more metaphoric and artistic stance. Collage is a good example of such clustering. General guidelines for clustering are: use a bottom‐up process of emergence; postpone early rationalisations and verbalisations; start grouping ideas on the basis of feeling and intuition; and use metaphoric names to identify clusters.