Comprehensive dataset of global innovation index panel data (2013–2022): Clustering with K-means and principal component analysis

Over the last decade, innovation has become a focal point for policymakers, business leaders, and researchers worldwide. In that context, this dataset draws on the annual Global Innovation Index (GII), compiled by the World Intellectual Property Organization (WIPO) and available on the World Bank's Prosperity Data360 portal, to offer a refreshed view of national innovation landscapes. It covers 118 economies with complete data from 2013 through 2022, organized into seven core pillars: Institutions, Human Capital and Research, Infrastructure, Market Sophistication, Business Sophistication, Knowledge and Technology Outputs, and Creative Outputs. Although the dataset also includes overall GII scores and Innovation Input/Output Sub-Indices for each year, those aggregated measures were not used in clustering or Principal Component Analysis (PCA) to preserve the detail of the seven pillars. To identify economies with similar innovation characteristics, we used the K-means algorithm, and the Elbow Method showed that five clusters worked best. We then applied this five-cluster framework in four different ways: focusing on input pillars only, focusing on output pillars only, combining all pillars, and using a version enhanced by Principal Component Analysis (PCA). PCA was introduced to reduce dimensionality and sharpen the divisions between clusters, which led to additional cluster labels for each scenario. Because it offers both breadth and depth in its indicators, this dataset can be especially helpful for those examining how different nations innovate, gauging where they stand in comparison to peers, or investigating longer-term trends. This dataset is particularly helpful for researcher, policymaker, or professional seeking solid data on innovation; the information here can inform strategic thinking and support evidence-based decision-making.