Applying deep learning to predict innovations in small and medium enterprises (SMEs): the dark side of knowledge management risk

Abstract

Purpose

This study aims to predict the dark side of knowledge management risk to innovation in Portuguese small and medium enterprises (SMEs). It examines the spinner innovation model factors of knowledge creation, knowledge transfer, private knowledge, public knowledge and innovation in uncertain environments.

Design/methodology/approach

The authors developed a conceptual model to support the analysis. The survey data stemmed from a sample of 208 Portuguese SMEs in Portugal. The authors analyzed the primary data from the ad hoc survey using the data mining (deep learning) technique.

Findings

The research sets out and tests factors relevant to understanding how to predict innovation in uncertain business environments. This study identifies four factors fostering innovation in SMEs: knowledge creation, knowledge transfer, public knowledge management and private knowledge management. Knowledge creation showed the best return and presented the closest relationship with innovation.

Originality/value

Innovation models generally measure the relationships between variables and their impacts on the economy (economic and regional development). Predictive models are considered in the literature as a gap to be filled, especially in an uncertain environment in the SME context.

Citation

Figueiredo, R., Ferreira, J.J., Camargo, M.E. and Dorokhov, O. (2023), "Applying deep learning to predict innovations in small and medium enterprises (SMEs): the dark side of knowledge management risk", VINE Journal of Information and Knowledge Management Systems, Vol. 53 No. 5, pp. 941-962. https://doi.org/10.1108/VJIKMS-09-2022-0294