Title: From 'big data' to 'smart data': algorithm for cross-evaluation as a novel method for large-scale survey analysis
Authors: Darko Kantoci; Emir Džanic; Marcel Bogers
Addresses: KanDar Enterprises Inc., 250 Commercial St., Suite 3005F, Manchester, NH 03101, USA ' Cambridge Innovative System Solutions Ltd., CPC1 - Capital Park, Fulbourn, Cambridge, CB21, 5XE, UK ' Unit for Innovation, Entrepreneurship and Management, Department of Food and Resource Economics, University of Copenhagen, Rolighedsvej 25, 1958 Frederiksberg C, Denmark
Abstract: Current research is increasingly relying on large data analysis to provide insights into trends and patterns across a variety of organisational and business contexts. Existing methods for large-scale data analysis do not fully capture some of the key challenges with data in large datasets, such as non-response rates or missing data. One method that does address these challenges is the SunCore algorithm for cross-evaluation (ACE). ACE provides a view of the whole dataset in a multidimensional mathematical space by performing consistency and cluster analysis to fill in the gaps, thereby illumining trends and patterns previously invisible within such datasets. This approach to data analysis meaningfully complements classical statistical approaches. We argue that the value of the ACE algorithm lies in turning 'big data' into 'smart data' by predicting gaps in large datasets. We illustrate the use of ACE in connection to a survey on employees' perception of the innovative ability within their company by looking at consistency and cluster analysis.
Keywords: statistical modelling; statistical algorithm; survey analysis; consistency analysis; cluster analysis; data trends; data patterns; data correlation; non-ignorable missing data; non-response missing data; cross evaluation; big data; smart data; innovation survey; food processing company.
International Journal of Transitions and Innovation Systems, 2018 Vol.6 No.1, pp.24 - 47
Received: 16 Jun 2017
Accepted: 21 Oct 2017
Published online: 14 Mar 2018 *