Pairwise comparison-based approach for predicting the ranking of global innovation capability

The Global Innovation Index (GII) is employed to rank the innovation capability for all countries and economies in the world, which was proposed by Cornell University, the INSEAD, and the World Intellectual Property Organization. To analyze the development status of innovation capability of a country or economy at any time, this paper proposes a pairwise comparison-based global innovation capability ranking prediction approach named PairwiseGIIRank. The experimental results show that: (1) The ranking of innovation capability can be predicted with PairwiseGIIRank effectively. (2) Compared with other classifiers, the approach performs the best when the k-nearest neighbors algorithm is used as its internal classifier. (3) To achieve the desired predictive effect, at least 5 indicators are needed. The ranking of innovation capability of countries and economies around the world can be predicted with PairwiseGIIRank before the GII report is released. It provides decision-making support for decision-makers to adjust their innovation capability development policies.

1 Introduction

Cornell University, INSEAND, and the World Intellectual Property Organization (WIPO) have released the Global Innovation Index (GII) report every year since 2007. The Innovation Index report provides a global ranking of national innovation capability for all countries and economies in the world, which gives the calculation method of GII, analysis of changes in the ranking of major countries, and the indicator systems used. The GII enables each country or economy to measure the gap between itself and other countries' innovation capability based on the scores and rankings.