Identifying industrial clusters with a novel big-data methodology: Are SIC codes (not) fit for purpose in the Internet age?

In this paper we propose using a novel big-data-mining methodology and the Internet as a new source of useful meta-data for industry classification. The proposed methodology can be utilised as a decision support system for identifying industrial clusters in almost real time in a specific geographic region, contributing to strategic co-operation and policy development for operations and supply chain management across organisational boundaries through big data analytics. Our theoretical discussion on discerning industrial activity of firms in geographical regions starts by highlighting the limitations of the Standard Industrial Classification (SIC) codes. This discussion is followed by the proposed methodology, which has three main steps revolving around web-based data collection, pre-processing and analysis, and reporting of clusters. We discuss each step in detail, presenting the experimental approaches tested. We apply our methodology to a regional case, in the North East of England, in order to demonstrate how such a big data decision support system/analytics can work in practice. Implications for theory, policy and practice are discussed, as well as potential avenues for further research.