Cluster analysis divides the data into distinct meaningful and useful groups that are often called clusters. If the goal of the researcher is to define and create meaningful clusters then it is important that the resulting cluster captures the natural structure of the data. For instance, cluster analysis has been often used for the purpose of grouping related document for the need of browsing, to identify those genes and proteins which have alike or similar functional traits or sometimes in the case of identifying groups of spatial locations that are prone to earthquakes. In most of the cases cluster analysis is the beginning stage for further analysis, for instance, compression of data. Whether for the purpose of understanding or utility, cluster analysis has been seen widely being used in different disciplines like psychology, social sciences, biology, and statistical research, identification of pattern, information retrieval and data mining.
Another perspective to cluster is innovation. They can also be viewed as reduced scale innovations. The outcome expected out of a cluster analysis is that the dynamics, characteristics of the system and the interdependencies are quite similar to that of a national innovation system, more importantly the cluster has a number of key advantages over the traditional sectorial perspective in trying to analyse innovation and the various innovative networks. The advantages are not restricted to just the process but also extend up to innovative policy making as well. The conclusive objective of clustering is to get rid of the imperfections within the innovative systems by the means of facilitating the efficient functions of these systems.
Interdependency is the main key to cluster analysis. But it is important for s researcher to remember that interdependency has many facets and it can be based or dependent upon trade linkages, innovation linkages, and the flow of knowledge or on the basis of common knowledge or common factor conditions. The focus primarily should be on the linkages and interdependencies between the actors in the value chain or the selected innovation system. One should not forget that these interdependencies are very relevant at different levels of the analysis and can be assessed by the use of various different techniques.